Volume 14 Issue 4

Embedded System-Based Characterization Of Solar Photovoltaic Panel Performance Across Varied Atmospheric Conditions

Authors: Siju George, Dr Ashish Kumar Rai

Abstract: This research paper presents an investigation into the application of smart embedded technology for the parametric analysis of solar photovoltaic (PV) panels. The proposed Arduino-enabled approach utilizes sensors to collect real-time data on various parameters such as temperature, irradiance, and voltage, allowing for comprehensive analysis of panel performance. Solar PV panels are widely used for harnessing renewable energy, but their performance can be affected by various factors such as shading, dust accumulation, and aging. Therefore, continuous monitoring and evaluation of PV panel parameters are crucial to ensure optimal energy generation.

Blockchain-Based Secure And Transparent Supply Chain Management Framework

Authors: V S S P L N Balaji Lanka, Dr. Rajidi Rammohan Reddy

Abstract: The growing complexities involved in the global supply chain system have brought about greater problems relating to data integrity, transparency, and trust issues. Traditional centralized systems of supply chain management are vulnerable to single point failure, lack of traceability, and fraudulent activities. This paper proposes a secure and transparent supply chain management system based on blockchain technology, utilizing smart contracts and cryptographic techniques to overcome these weaknesses. The proposed framework is a combination of IoT-based data ingestion with a blockchain network, using optimal queuing technique for efficient transaction processing. The simulation results reveal considerable gains in terms of trace-back capability (decrease from 95 seconds to 8 seconds), transactional efficiency (120 transactions per minute), and data integrity (99.9% fraud detection capability). Comparison with the traditional framework reveals a 14-times gain in dispute resolution and 70% improvement in trust indicators,

DOI: http://doi.org/10.5281/zenodo.21338046

Deep Learning-Based Student Performance Prediction And Analysis

Authors: Lavanya CM, M. Karthiyayini

Abstract: Predicting student performance accurately is vital in early detection of high-risk students and intervention in academic activities in higher education. This paper provides a systematic review on deep learning techniques used in student performance prediction. Recent developments in deep learning techniques such as hybrid stacked ensemble model, Gated Long Short-Term Memory network, Bidirectional LSTM with SHAP interpretability and integrated feature transformer networks have been discussed. The results of the study indicate that ensemble models utilizing several deep learning algorithms with performance-based weight give remarkable performance, with recall rate of 98.26%, precision rate of 99.51%, and F1-score of 98.88% in identifying at-risk students. Deep Learning based Gated LSTM models which use Dove optimization method gives 98.85% classification accuracy. Hybrid Deep Learning models, which integrate time-based behavioral patterns with static attributes of students give impressive results with various educational datasets. It has been shown that Deep Learning models provide better results compared to machine learning algorithms with accuracy rates over 97%. Data imbalance, lack of interpretability and generalization of educational data pose challenges.

DOI: http://doi.org/10.5281/zenodo.21338215

Hybrid Graph Neural And Machine Learning Architecture For Complex Network Intelligence

Authors: Roshan Rukshana Sulaima Lebbe, Padmaja C

Abstract: The growing sophistication in the structure and behavior of current networked systems requires sophisticated methods that can effectively uncover and understand the complicated patterns and relationships in graph-based data. In this paper, a new hybrid approach is introduced that utilizes the combination of Graph Neural Networks (GNNs) along with conventional machine learning techniques to improve complex network intelligence. This proposed method uses Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), and Graph Autoencoders (GAEs) to develop advanced node embeddings, which are useful for understanding both local and global graph structures along with using ensemble learning algorithms for decision making. It is observed through quantitative analyses on several benchmark datasets that the hybrid architecture performs better than the individual methods in terms of accuracy for node classification, anomaly detection, and influential nodes identification tasks.

DOI: http://doi.org/10.5281/zenodo.21349178

IoT-Based Multi-Parameter Stroke Risk Monitoring System Using Wearable Sensors And Machine Learning

Authors: G Thamarai Selvan

Abstract: Stroke is a serious medical condition that often leads to longterm disability or even death if not identified early. Continuous monitoring of vital health parameters can help detect early warning signs and reduce the severity of such conditions. In this study, a smart healthcare system is designed using Internet of Things (IoT) technology with ai to monitor multiple physiological signals in real time. The system uses wearable sensors to measure heart rate, blood oxygen level (SpO₂), body temperature, and body movement. These values are collected using an ESP32 microcontroller and sent to a cloud platform for storage and monitoring. A machine learning model based on the Random Forest algorithm is used to analyze the collected data and classify the risk level into safe, moderate, or high. In addition, the system provides immediate alerts using a buzzer, LED indicators, and mobile notifications whenever abnormal readings are detected. The results show that the model performs with good accuracy and can help in early identification of stroke-related risks. This system is affordable, easy to use, and suitable for remote healthcare monitoring.

DOI: http://doi.org/10.5281/zenodo.21351006

Extraction Of The Kinetic Freeze-Out Temperature By An Alternative Method : A Fokker–Planck Analysis

Authors: Hassan Ali Khan, Hadiqa Qadir

Abstract: The extracted kinetic freezeout temperature reflects the thermal conditions of the system at the final stage of elastic interactions, and is typically obtained through simultaneous fits of transverse momentum spectra using blast-wave-type models that also yield the average radial flow velocity. In this study, we extract the kinetic freeze-out temperature, Tkin, from the transverse momentum spectra of identified negativley charged hadrons π−, K− and p− produced in Au–Au collisions across the RHIC Beam Energy Scan, spanning √sNN = 7.7 to 200 GeV. Our analysis builds upon effective temperatures obtained in our previous work, from which we isolate the kinetic decoupling temperature by applying a linear fit, Teff = m m0 + Tkin, to the particle mass dependence of the effective temperatures. We present two independent sets of fitting results, which show consistent qualitative behaviors and provide robust estimates of the freeze-out parameters. Our extracted values of Tkin reveal two clear and systematic trends. For any fixed collision energy, the freeze-out temperature decreases monotonically as we move from central to peripheral collisions, reflecting the diminishing system size, lower energy density, and reduced rescattering in the later stages of the fireball evolution. For a fixed centrality bin, the temperature rises with increasing beam energy, but the rise is not uniform: a steep increase at low energies is followed by a plateau around 19.6–39 GeV, which then gives way to a renewed rise at the highest RHIC energies. This non-monotonic behaviour is interpreted as evidence for a change in the underlying degrees of freedom, consistent with the system transitioning from a baryon-rich hadronic phase through a possible crossover region and into a parton-dominated phase at the highest energies. The results demonstrate that the kinetic freeze-out temperature is a sensitive probe of the system size, initial energy density, and the stiffness of the equation of state. Our findings provide important constraints for hydrodynamic models and highlight the utility of the Fokker–Planck approach in extracting freeze-out parameters from experimental data.

DOI: http://doi.org/10.5281/zenodo.21356234

Structural Performance Evaluation Of A Five-Storey RCC Hostel Building Under Gravity And Seismic Loads

Authors: Ahmed Raza, Zaheer Ahmed, Sarfaraz Malik, Hasnain Ahmed, Muhammad Azib

Abstract: The rapid growth in university enrollment has increased the demand for safe and economical student accommodation, particularly in seismically active regions. This study presents the structural modeling, analysis, and design of a five-storey reinforced concrete (RCC) hostel building located in Rahim Yar Khan, Pakistan, using ETABS software. The structure was designed in accordance with ASCE 7-16 for loading and ACI 318-14 for reinforced concrete design. Gravity loads (dead, super dead, and live loads) and lateral loads (seismic loads in X and Y directions) were applied to evaluate structural performance. Results show that the maximum story displacement occurred at the roof level with a value of 0.3299 inches, which is within permissible limits. The maximum story drift was recorded at Story 4 with a value of 0.000553, significantly lower than the allowable drift limit of 0.020 prescribed by ASCE 7-16. The highest story stiffness of 3818.72 kip/in was observed at the first storey, decreasing gradually towards the top. Base reaction results indicated a seismic base shear of -96.119 kip in the X-direction and overturning moments of – 4172.55 kip-ft and 3649.15 kip-ft about the Y and Z axes, respectively. These results confirm that the proposed RCC structural system satisfies strength, stiffness, and serviceability requirements, making it suitable for long-term student accommodation.

DOI: http://doi.org/10.5281/zenodo.21358437

EFL Learners’ Perceptions of the Role of Commercial Game-based Interaction in L2 Learning

Authors: Ali Mohammad

Abstract: This study is an endeavour to elucidate the pedagogic value of interaction within video game environments by the calculation of selected learning opportunities (negotiation of meaning, language-related episodes, feedback, noticing) by using descriptive statistics and learner perception of seven English language learners. The study adopted one cognitive approach to address the first topic and two sociolinguistic approaches to account for the latter. Such collaborations were found to stem from processes of community of practice, language socialisation and willingness to communicate. Additionally, the thematic analysis of the narratives queries suggests wide acceptance of the use of video games towards L2 learning. The findings advocate for the role of the interaction in video games, while also paying attention to numerous factors which affect it: personality, feedback, region and communicative competence. In the study, accompanied with the findings are a set of recommendations. This provides a new approach for EFL educators who can blend leisure games with formal educational setting.

DOI: https://doi.org/10.5281/zenodo.21369718

Generative AI in Software Development

Authors: Dhanushree K R, Aniket Singh, Associate Professor Dr. V. Sathya

Abstract: Generative AI is transforming software development through code generation, testing, debugging, documentation, and project support. This review examines the evolution, applications, benefits, challenges, and future scope of Generative AI in software engineering.Generative Artificial Intelligence (Generative AI) has emerged as a revolutionary technology that is transforming software development by automating coding tasks, improving productivity, and supporting developers throughout the Software Development Life Cycle (SDLC). Powered by Large Language Models (LLMs), Generative AI enables intelligent code generation, debugging, software testing, documentation, and code review, thereby reducing development time and minimizing human effort. This review paper presents a comprehensive analysis of the role of Generative AI in modern software engineering by examining recent research, industrial applications, and technological advancements. The paper discusses the adoption of AI-powered tools such as GitHub Copilot, ChatGPT, Amazon CodeWhisperer, and Google Gemini, highlighting their contributions to improving software quality and developer efficiency. Furthermore, it explores the key benefits of Generative AI, including enhanced productivity, faster development cycles, improved collaboration, and support for novice programmers. Despite these advantages, several challenges remain, such as data privacy concerns, security vulnerabilities, inaccurate code generation, copyright issues, and ethical considerations associated with AI-generated content. The paper also examines future research directions, including autonomous software engineering, AI-assisted DevOps, personalized coding assistants, and responsible AI governance. Based on an extensive review of recent literature, this study concludes that Generative AI has significant potential to reshape the future of software development when combined with human expertise and robust ethical practices. The findings provide valuable insights for researchers, educators, software professionals, and students interested in understanding the evolving impact of Generative AI on software engineering.

CognitoNavigo: EEG-Based Maze Game

Authors: Atharva Koli

Abstract: This paper presents CognitoNavigo, an EEG- based maze game that leverages Brain- Computer Interface (BCI) technology for interactive navigation. The system converts real- time EEG signals into control commands, allowing users to navigate a physical maze using their brain activity. The proposed approach involves EEG signal acquisition, feature extraction, and classification for interpreting user intentions. A distinctive maze design was developed using SolidWorks, featuring an innovative structure that enhances user engagement. The design ensures optimal placement of actuators and finalizing the ergonomic layout for user accessibility, and ideal integration of mechanical and electronic components. Detailed simulations and assembly models contribute to enhancing the system’s structural and functional effectiveness. Experiments trials demonstrated a 60% success rate in command recognition, with corresponding motor actuation. A threshold- based signal processing system proved crucial for reliable command interpretation. Notably, the results highlight the importance of adaptive user training, as participators progressively enhanced their control accuracy over time. The system showcases potential operations in cognitive training, neurorehabilitation, and brain- controlled gaming by effectively bridging human cognitive processes with physical actuation.

Effect of Surya Namaskar, Tadasana, Vriksasana, Trikonasana, Bhujangasana, Virabhadrasana, Tulasana, Matsyadrasana, Garudasana, Gaomukhasana, Uttan Padasana, Bhramari Pranayam, Shitali Pranayam, Anulom-Vilom Pranayam, Nadi Sodhan Prayanam, Kapalbhati, Meditation on Intraocular Pressure and Tear Film Evaluation

Authors: Bhumika Yadav, Associate Professor Dr Gaurav Kumar Bhardwaj, Assistant Professor Mr Animesh Mondal

Abstract: Purpose: This study examines the impact of different yoga asanas and pranayama practices on intraocular pressure (IOP) and tear film stability. This research aims to assess the ocular health advantages of yoga by examining physiological factors related to eye pressure control and tear film stability. Methods: A cross-sectional study was performed on 40 persons aged 20 to 40 who consistently engaged in yoga practice. The participants completed pre- and post-exercise evaluations utilizing the Non-Contact Tonometer to determine intraocular pressure (IOP) and employed both the Tear Break-Up Test (TBUT) to analyze tear film stability. The yoga practice spanned 40-50 minutes and comprised asanas including Surya Namaskar, Tadasana, Vriksasana, Trikonasana, and Bhujangasana, along with breathing techniques such as Bhramari Pranayama, Anulom-Vilom, and Kapalbhati. The data analysis was to evaluate yoga's effect on ocular metrics. Results: Surya Namaskar markedly affected intraocular pressure and tear film stability, demonstrating a statistically significant difference relative to other yoga asanas. Tadasana, Vriksasana, Bhujangasana, and Virabhadrasana demonstrated minimal impact on intraocular pressure (IOP), whereas Kapalbhati resulted in a temporary elevation in IOP due to vigorous exhalation. Pranayama techniques, including Anulom-Vilom and Nadi Sodhan, exhibited a progressive reduction in intraocular pressure over time. Most yoga poses did not substantially affect tear film integrity, except Surya Namaskar, which showed a considerable influence. Conclusion: The research emphasizes Surya Namaskar as a potentially advantageous yoga practice for affecting intraocular pressure and tear film stability. Other yoga asanas have shown negligible effects, however, pranayama approaches demonstrated potential for long-term intraocular pressure stability. Additional research with expanded sample size is advised to corroborate these findings and investigate the therapeutic ramifications of yoga on eye health.

Morphometric Analysis Of Vandaman Eru Watershed, Southern India Using Remote Sensing And GIS

Authors: Gara Raja Rao, Sangaraju Siddi Raju, K. Satyanarayananaidu, Kambam Swetha, Yenda Padmini, Akkupalli Surekha, Mallula Srinivasa Rao

Abstract: Morphometric analysis is a crucial part of the complete evaluation of Morphometric analysis are significant processes in hydrology, landform evolution and environmental processes at drainage basin scale. The morphometric parameter of stream, area and relief aspects obtained through the integrated remote sensing (RS) and geographic information system (GIS) techniques will helping to understand the hydrological behaviour, erosional vulnerability and groundwater recharge potential of a u-shaped Vandaman Eru Watershed (508.98 km2). This watershed is semi-arid, and we can observe a branching drainage pattern that relates to lithological homogeneity with little tectonic distortion. This study revealed the significant spatial variability in drainage density throughout the watershed. The watershed-integrated drainage density is calculated on the basis of total stream length (1,855.85 km) compared with total basin area (508.98 km2) computes at 3.65 km/km2 which indicates fine dissection and a high surface runoff generation capacity that is characteristically found in higher-altitude, peripheral terrain. Drainage density in the context of geomorphic region is 0.23 to 0.69 km/km2 for allu-vial plains landform with lowest dissection and high infiltration potential at sub-basin level compared to other geomorphic units. This dichotomy of low sub-basin drainage densities versus high basin-scale drainage density is indicative of the criti-cal spatial heterogeneity of the watershed, one which will be explicitly differentiated throughout this manuscript. Important key linear, areal and relief morphometric parameters extracted for understanding channel development mechanism, basin geometrics, runoff potential and erosion susceptibility. The results demonstrated a strong inverse exponential correlation (R2 = 0.81) between the drainage density and infiltration rate within zones defined spatially. Establishing the hydrological segregation controlled by spatial variability in drainage density across the watershed via a holistic morphometric analysis and identifying valley fills and pediplain zones as regional-focused target sites for artificial groundwater recharge.

Applications of YOLO in the Oil and Gas Industry

Authors: Ahmad Alotaibi, Masnour Asiri, Naif Qahtani, Abdullah Aldawsari

Abstract: The You Only Look Once (YOLO) algorithm has become one of the most widely used deep learning frameworks for real-time object detection due to its high detection accuracy, computational efficiency, and ability to simultaneously perform object localization and classification. This paper reviews the applications of YOLO in industrial environments, with particular emphasis on battery leakage detection, facial detection for plant access control, and hazardous gas leakage detection. These applications demonstrate how YOLO enhances manufacturing quality, strengthens industrial security, improves workplace safety, and supports intelligent automation through rapid and reliable object detection. Despite its significant advantages, the practical implementation of YOLO presents several challenges. The model requires substantial computational resources and specialized hardware to achieve real-time performance, particularly when processing high-resolution images or multiple video streams. Furthermore, YOLO relies heavily on large, diverse, and accurately annotated datasets, while class imbalance in industrial datasets can reduce detection performance for rare but safety-critical objects. To overcome these limitations, several solutions are discussed, including lightweight YOLO architectures, model compression techniques, edge artificial intelligence (AI) hardware acceleration, data augmentation, synthetic data generation, balanced sampling methods, and advanced loss functions. Overall, the continuous evolution of YOLO architectures and optimization techniques has significantly improved its applicability across industrial sectors, making YOLO a promising and adaptable solution for smart manufacturing, industrial inspection, surveillance, and safety monitoring within Industry 4.0 environments.

DOI: http://doi.org/10.5281/zenodo.21406036

Evaluating The Aerodynamic Impact Of Wing- Horizontal Tail Interference On The Su-27 Via ANSYS

Authors: Van Viet Vo

Abstract: This paper investigates the aerodynamic interference of the wing-horizontal tail combination on the Su-27 using ANSYS. Full-scale numerical simulations at a 10 km altitude reveal that interference occurs exclusively at negative angles of attack, depending on flight speed. Here, wing airflow alters the flow field over a portion of the horizontal tail, modifying the aircraft's lift, drag, and aerodynamic efficiency. This study resolves high-precision computation challenges by identifying the specific tail segment affected by the downwash angle, enabling accurate lift and drag coefficient calculations to reliably serve research and flight training.

DOI: http://doi.org/10.5281/zenodo.21413403

Real-Time Speech-to-Text And Speaker Diarization Using Whisper And Pyannote Embeddings

Authors: Shivani Chauhan, Rimmy, Ashish Prajapati

Abstract: Speech-to-text (STT) and multi-speaker diarization have emerged as crucial components in intelligent communication systems, virtual meeting platforms, assistive technologies, and large-scale multimedia analytics. Recent advancements in transformer-based architectures such as Whisper and self-supervised pipelines like Pyannote have significantly improved transcription quality and speaker discrimination, enabling highly accurate multi-speaker processing even on consumer hardware. A system that integrates Whisper for multilingual transcription, Pyannote for diarization, and a new speaker identity recognition module that uses voice embedding is presented in this research as a modular, scalable, and real-time integrated system. Real-time microphone input, audio uploads, and live transcript visualization are all supported by a full-stack web platform that uses React and Flask. The Word Error Rate (WER) of 6.2% and the Diarization Error Rate (DER) of 13% were observed in experiments conducted on 5 hours of multi-speaker audio. 87% identity accuracy is achieved by personalized speaker recognition. Practical usability is demonstrated by the proposed system for meetings, lectures, podcasts, and automated captioning.

DOI: http://doi.org/10.5281/zenodo.21418016

Text Summarization For News Articles Using An AI Chatbot

Authors: Dr. Brij Mohan, Sagar Chaudhary

Abstract: Massive amounts of textual content every day are produced due to the rapid expansion of digital news channels. Quickly drawing significant conclusions from lengthy news articles may be a challenge for readers, journalists, and academics. The problem of condensing long articles into brief and instructive summaries has been solved by automating text summarizing, which preserves the fundamental meaning and context. The design and implementation of an AI chatbot-based system for automatic text summarization of news articles are presented in this research paper using deep learning and NLP approaches. The suggested system provides support for abstract, hybrid, and extractive summarization techniques. Despite the sophistication of transformer-based models like BERT[4, T5[7], and PEGASUS[7] Abstract summarizing is done by using traditional extractive techniques like TF- to provide summaries that are human-like. The statistical significance of IDF[2] and TextRank[8] is used to identify significant phrases. The modular system architecture includes preprocessing, summary engine, evaluation module, and interactive chatbot interface, making it easy for users to enter news items and receive summarized results instantly.

DOI: http://doi.org/10.5281/zenodo.21418085

Inbox Ai – Personalized Email Generator with Tone Selection

Authors: Dr Himanshu Tyagi, Anurag Chandna, Deepak Bhatt

Abstract: Writing emails with correct tone is a common challenge for students, professionals and businesses. Many times the message is correct, but the tone does not match the context — resulting in misunderstandings, poor communication or unprofessional impact. Hence there is a need for a system that can automatically generate a well-structured email and adjust the tone according to the requirement. In this paper we propose a Personalized Email Generator with Tone Selection, which uses AI/NLP models to convert raw user input (idea/points) into a complete professional email. The system provides multiple tone options such as Formal, Friendly, Professional, Casual, Direct, and Promotional, and instantly rewrites the text accordingly. This helps users generate consistent, context-correct communication without manual editing or writing skills. The system can be deployed as a web app, extension or integrated tool for day-to-day email writing.

DOI: http://doi.org/10.5281/zenodo.21418189

Object Detection In Autonomous Vehicles

Authors: Rohan Chaudhary, Mayank Chauhan, Dr. Partap Singh

Abstract: In the context of autonomous vehicle (AV) perception, this study examines the implementation and performance assessment of the You Only Look Once version 5 (YOLOv5) model for object detection (OD). Safe and reliable navigation requires high-throughput algorithms, which require accurate and real-time detection. Roboflow's carefully selected dataset of approximately 15,000 photos was used in the study, with a focus on specific classes like cars, trucks, pedestrians, traffic lights, and bikes. PyTorch was used to train the YOLOv5 architecture for 100 epochs to enhance generalization across different lighting and angle conditions, and robust data augmentation techniques like brightness adjustment and scaling were included. The model's efficiency is attributed to its single-stage detection approach, which places emphasis on the necessary inference speed for real-time decision-making in autonomous systems. The trained model achieved a mean average precision (textmAP) of around 85%, and both precision and recall metrics exceeded 80% in terms of numbers. Critical analysis revealed that detection accuracy was significantly higher for large, frequently occurring classes (cars) than for smaller, safety-critical classes (pedestrians and traffic lights), indicating a performance gap. Further architectural improvements are required due to the structural limitation caused by class reduction and scale variations. The paper concludes with a roadmap for improved robustness, which includes migrating to sophisticated architectures like YOLOv8 or Detection Transformers (DETR), to achieve near-perfect recall rates for high-consistency objects.

DOI: http://doi.org/10.5281/zenodo.21418279

Skin Cancer Detection Using CNN

Authors: Parul Taygi, Neetu Mourya, Nisha sharma

Abstract: Skin cancer is one of the most common forms of cancer globally, with melanoma posing life-threatening risks if not detected early. Traditional diagnosis relies on visual clinical examination, dermoscopy, and biopsy, all of which require expert dermatologists and are subject to human interpretation. Recent advancements in Artificial Intelligence, especially Convolutional Neural Networks (CNNs), have enabled highly accurate automated analysis of dermoscopic images. This research presents a deep learning–based skin cancer detection system that classifies dermoscopic images into benign or malignant categories. The system uses standardized image preprocessing, augmentation, and a custom CNN architecture trained on publicly available datasets such as ISIC and HAM10000. Performance metrics, including accuracy, precision, recall, F1-score, demonstrate that CNNs can effectively extract hierarchical skin lesion features. Experimental results show high diagnostic performance comparable to dermatologists under controlled settings. The model has potential applications in telemedicine, early screening, and decision-support systems for dermatologists.

DOI: http://doi.org/10.5281/zenodo.21418363

A Personalized Fitness Application For Injury Rehabilitation Using Artificial Neural Networks

Authors: Gautam Tyagi, Deepak Saini, Bhanu Partap

Abstract: This paper presents a personalized fitness mobile application with the use of Artificial Intelligence (AI), namely Artificial Neural Networks (ANN), which contributes to injury rehabilitation. It automatically generates customized workout plans on a weekly basis, taking into consideration user-supplied data on issues such as injury type, fitness level, body mass index, and available equipment. The system continuously updates the workout plans with the help of feedback provided by the user to achieve the best recovery process. Built with React Native for the frontend and Django for the backend, this application guarantees access anytime, anywhere, with smooth data treatment. The paper further reflects on the general implications of using AI in e-fitness, stresses intuitive user interfaces, and furthers possibilities such technologies could undergo.

DOI: http://doi.org/10.5281/zenodo.21418419

Secure Authentication Via Facial Recognition: A Research Review & Implementation Study

Authors: Dr MD Iqbal, Dr Partap, Vipin kumar dhiman

Abstract: Facial recognition technology has gained significant attention as a biometric authentication mechanism due to its efficiency, user convenience, and non-intrusive characteristics. With the rapid expansion of digital platforms in sectors such as banking, mobile computing, workplace security, and online services, the need for robust and secure authentication systems has become increasingly critical. This research paper explores facial recognition as a secure authentication technique by examining its underlying working principles, security advantages, potential vulnerabilities, and associated ethical concerns. A qualitative, review-based methodology is adopted, involving an extensive analysis of existing scholarly literature, industry reports, and real-world applications, including smartphone authentication and enterprise-level security systems. The findings indicate that facial recognition enhances usability and reduces reliance on traditional password-based systems; however, it remains susceptible to threats such as spoofing attacks, deep fake manipulation, data breaches, and algorithmic bias. Additionally, privacy issues and regulatory constraints pose challenges to its widespread adoption. The study concludes that facial recognition can serve as a reliable authentication method. When supported by live ness detection techniques, encryption mechanisms, and stringent data protection policies. Future research should focus on bias reduction, advanced anti-spoofing solutions, and privacy-preserving biometric frameworksy.

DOI: http://doi.org/10.5281/zenodo.21425027

Optimizing Diabetes Prediction Accuracy Via Integrated Random Forest, SVM, And Logistic Regression

Authors: Dr Chunnu Lal, Dr Satender Kumar, Dr Raj kumar

Abstract: The prevalence of Diabetes Mellitus, a chronic metabolic disorder, is rapidly increasing worldwide, which is creating an urgent need for early risk assessment tools that are fast, accessible, and accurate. This research presents a comprehensive study of a diabetes prediction web application that combines three heterogeneous machine learning classifiers—Random Forest, Support Vector Machine (SVM), and Logistic Regression—in a soft-voting ensemble architecture. The Pima Indians Diabetes Database (PIDD) was used to train and validate the integrated system, which uses eight clinically relevant health parameters for binary classification. The ensemble model's test set accuracy was 81.04%, which is a significant improvement over the individual base classifiers (LR: 76.60%, SVM: 75.32%, RF: 78.45%). The system demonstrated strong clinical utility with a precision of 77.6% and a recall rate of 65.0%, establishing a critical balance for medical screening applications. By deploying the complete architecture through a lightweight Flask-based web interface, it was possible to predict risks in real-time with millisecond-level inference latency. This work validates the effectiveness of soft-voting ensemble methodologies in achieving robust classification accuracy for high-stakes healthcare applications and demonstrates a scalable, practical implementation pattern for deploying machine learning models to address significant public health challenges.

DOI: http://doi.org/10.5281/zenodo.21425052

NLP-Based Sentiment Analysis Framework For Education Industry 4.0

Authors: Raj Kumar, Monti Saini, Shilpy Sharma

Abstract: Sentiment analysis is a powerful computational technique used to identify, extract, and interpret subjective information embedded within textual data. It focuses on recognizing and classifying opinions, attitudes, emotions, beliefs, and feelings expressed by individuals through written language. In the educational domain, sentiment analysis plays a crucial role in understanding student behaviour, engagement, motivation, and overall learning experiences. By analysing linguistic features such as word choice, sentence structure, contextual meaning, and sentiment lexicons, sentiment analysis systems can detect a wide range of emotional states, including positive, negative, neutral, and subtle affective expressions. Recent advancements in Natural Language Processing (NLP) have significantly improved sentiment analysis performance through the adoption of transformer-based architectures and Large Language Models (LLMs). These models are trained on massive datasets using deep neural networks, enabling them to capture contextual and semantic nuances more effectively than traditional machine learning approaches. Models such as BERT, RoBERTa, and ELECTRA have demonstrated remarkable success across various NLP tasks, including sentiment classification and emotion detection. However, the practical deployment of such models in educational settings often faces challenges related to limited labelled data, high computational costs, and resource constraints. To address these challenges, this project investigates the application of LLMs for student sentiment analysis within an Education 4.0 framework by examining three primary adaptation strategies: zero-shot learning, N-shot learning, and fine-tuning approaches. The proposed system analyses student feedback to assess sentiment polarity, emotional dimensions, and a composite Learning Quotient (LQ) that reflects engagement, comprehension, motivation, collaboration, and critical thinking. Experimental observations indicate that different adaptation strategies yield significantly varied performance outcomes, highlighting the importance of selecting appropriate modelling techniques based on available resources and application requirements. Overall, the results emphasize the strong potential of LLM-based sentiment analysis systems in enhancing data-driven decision-making, personalized learning, and adaptive educational environments despite existing resource limitations.

DOI: http://doi.org/10.5281/zenodo.21425087

Towards Intelligent Prediction Of Critical Process Died Errors In Windows Systems: A Comprehensive Review Of AI-Driven Detection And Prevention Techniques

Authors: Mr. Harunmiya Sirajmiya Malek

Abstract: One of the most difficult problems affecting the dependability and security of contemporary Windows operating systems is still kernel-level failures. The Critical Process Died error is the most important of these errors since it causes a Blue Screen of Death (BSOD) when a crucial operating system process fails, abruptly ending system execution. In addition to disrupting regular computer operations, these failures also affect cloud platforms, enterprise services, AI workloads, and digital security infrastructures. Recent developments in AI-driven system monitoring have shown a great deal of promise for spotting unusual system behavior before disastrous catastrophes take place. Recent studies on log-based anomaly detection, graph neural networks, transformer-based language models, contrastive learning techniques, AIOps frameworks, and eBPF-enabled observability for predictive system monitoring (2022–2024) are all critically examined in this study. The study also evaluates current methods according to their capacity for detection, computational effectiveness, scalability, and suitability for Windows-based settings. In order to determine future paths for intelligent operating system reliability, current research trends, constraints, and unresolved issues are examined. Lastly, a hybrid AI-driven framework that incorporates anomaly detection, kernel telemetry, and log analytics is suggested to strengthen the resilience of next-generation computing systems and improve early failure prediction.

DOI: http://doi.org/10.5281/zenodo.21426385

Short Term Electricity Price Forecasting Using Hybrid Machine Learning And Feature Selection Techniques

Authors: Manjesh kumar, Dr. Jaya Shukla, Dr. Rajnish Bhasker

Abstract: Short-term electricity price forecasting (STEPF) plays a crucial role in modern deregulated electricity markets by enabling effective energy trading, demand-side management, grid stability, and operational planning. However, the highly volatile and nonlinear behavior of electricity prices, influenced by factors such as electricity demand, renewable energy integration, fuel prices, weather conditions, and market uncertainties, makes accurate forecasting a challenging task. To address these challenges, this paper proposes a novel hybrid Machine Learning framework that integrates advanced data preprocessing, adaptive feature selection, multi-stage deep neural learning, residual error correction, and explainable artificial intelligence (XAI) into a unified forecasting architecture. Initially, missing values are estimated, outliers are removed, time-series decomposition is performed, and normalized sliding-window sequences are generated to improve data quality. A hybrid feature selection strategy combining Mutual Information, ReliefF, and Adaptive Grey Wolf Optimization identifies the most informative features while reducing redundancy and computational complexity. The selected features are then processed through a hybrid CNN–BiLSTM–Temporal Attention–Transformer network to capture local spatial characteristics, long-term temporal dependencies, and global contextual relationships. Furthermore, a lightweight Extreme Learning Machine-based residual correction module refines the initial predictions by minimizing residual forecasting errors. To enhance model transparency, SHAP and Integrated Gradients are employed to interpret feature contributions and prediction behavior. Experimental analysis conducted on a realistic synthetic electricity market dataset demonstrates that the proposed framework achieves superior forecasting performance with an MAE of 0.48, RMSE of 0.67, MAPE of 1.21%, and an R² score of 0.995, outperforming conventional machine learning and Machine Learning benchmark models while maintaining robust generalization and computational efficiency.

Sentiment Analysis of Social Media Text for Identifying Public Opinion, Trends, and Consumer Behavior

Authors: Ms. Ruchika Kadu, Ms. Vaishnavi Nawle, Mr. Jayesh Bisane, Mr. Nikhil Barapatre

Abstract: The exponential growth of social media platforms has fundamentally transformed the way individuals communicate, share opinions, express emotions, and influence public discussions across the world. Platforms such as X (formerly Twitter), Facebook, Instagram, Reddit, LinkedIn, and YouTube generate billions of user-generated posts, comments, reviews, and discussions every day. These digital interactions represent a valuable source of information that reflects people's attitudes, emotions, preferences, experiences, and behavioral patterns regarding products, services, political events, healthcare, education, entertainment, and social issues. As organizations increasingly rely on data-driven decision-making, extracting meaningful knowledge from this vast amount of unstructured textual data has become an important research area. Sentiment Analysis, also known as Opinion Mining, has emerged as one of the most effective Artificial Intelligence (AI) techniques for automatically identifying and classifying emotions, opinions, and attitudes expressed in textual content. Sentiment Analysis integrates Natural Language Processing (NLP), Machine Learning (ML), Deep Learning (DL), and computational linguistics to analyze textual information and determine whether a particular opinion is positive, negative, or neutral. Unlike traditional data analysis methods that mainly focus on structured numerical information, sentiment analysis enables organizations to interpret human emotions and understand public perceptions from unstructured social media content. This capability provides significant advantages in understanding customer satisfaction, monitoring brand reputation, identifying emerging trends, predicting market behavior, and supporting strategic decision-making. The increasing adoption of Artificial Intelligence has further enhanced the accuracy and scalability of sentiment analysis systems, making them capable of processing millions of social media posts in real time.

DOI: https://doi.org/10.5281/zenodo.21428007

Integrated Battery Supercapacitor System for Optimized Solar Traction Drives

Authors: Ayush Kumar Yadav, Assistant Professor Anurag Singh

Abstract: The increasing demand for sustainable transportation has accelerated the development of solar-powered electric vehicles; however, the inherent intermittency of solar energy and rapid variations in traction load pose significant challenges in maintaining system stability and efficiency. This paper presents the design and simulation of a battery–supercapacitor hybrid energy storage system for improving the dynamic performance of a solar electric vehicle. A novel Adaptive Predictive Droop Control strategy is proposed, which integrates multi-timescale power decomposition with adaptive droop characteristics and predictive correction of load and solar variations. The proposed control dynamically distributes power between the battery and supercapacitor based on system conditions, thereby reducing battery stress and enhancing transient response. A detailed mathematical model of the system components, including battery, supercapacitor, and DC bus dynamics, is developed and implemented in MATLAB. The system is evaluated under multiple operating conditions such as load variations, acceleration, regenerative braking, and solar fluctuations. Simulation results demonstrate improved DC bus voltage stability, reduced battery current peaks, and enhanced system efficiency compared to conventional battery-only systems. The proposed method provides a reliable and efficient solution for next-generation solar electric vehicle energy management.

Predictive Analytics for Stock Markets Using Machine Learning

Authors: Akshatha N S, Keerthi T S, Pallavi B, Associate Professor Dr. Venkatesh

Abstract: Stock price forecasting is a popular and important topic in financial and academic studies. Share market is an volatile place for predicting since there are no significant rules to estimate or predict the price of a share in the share market. Many methods like technical analysis, fundamental analysis, time series analysis and statistical analysis etc. Since stock trading is so important to the financial industry, investors are constantly looking for trustworthy strategies to predict market fluctuations. Predicting the future values of stocks or other financial assets that are traded on exchanges is known as stock market prediction. The use of machine learning (ML) in stock price forecasting is investigated in this paper. Techniques including time-series forecasting, technical analysis, and fundamental analysis are commonly used by traders and investors to inform their investment decisions. The main programming language used to create the machine learning models in this study is Python. The suggested method trains an ML model using historical stock data with the goal of finding trends and producing predictions based on insights gleaned from the data. The Support Vector Machine (SVM) method is specifically used in the study to forecast stock values under a range of market scenarios. Both large-cap and small-cap stocks are used to test the model's performance and daily and real-time data are used to examine price changes. The goal of this research is to improve stock price prediction accuracy and give investors useful decision-making assistance by incorporating machine learning techniques.

DOI: https://doi.org/10.5281/zenodo.21454555

News Classification Using Machine Learning And Deep Learning: A Comparative Study

Authors: Dr. Satender, Dr. Brij Mohan, Dr. Raj Kumar

Abstract: In Natural Language Processing (NLP), news classification is a key task that involves automatic categorization of news articles into predefined topics, which allows for efficient content organization and information retrieval. This paper presents a news classification model that is based on machine learning. Collecting news datasets, pre-processing text, extracting features using TF-IDF, and classification using multiple algorithms, including Naive Bayes, Logistic Regression, and Deep Learning (LSTM), are the main tasks of the proposed method. Technology, Business, Sports, Education, and Entertainment are among the categories included in the dataset. The model's evaluation is based on metrics such as accuracy, precision, recall, and F1-score. The LSTM model has experimental results that demonstrate its highest accuracy of 91%, surpassing Logistic Regression (87%) and Naive Bayes (85%). Automated content categorization in the media sector is furthered by the study.

DOI: http://doi.org/10.5281/zenodo.21455605

A Research Review & Implementation Study Students Performance Prediction

Authors: Dr Tanu Gupta, Dr Mridula, Jaishree Goyal

Abstract: Small daily factors, such as attendance, study effort, family support, sleep, and co-curricular engagement, have a significant impact on student achievement. Often, these signals are only apparent after a student's grades drop, which makes timely intervention difficult. The purpose of this paper is to present a Student Performance Prediction System that combines an interpretable machine-learning pipeline with a web-based interface to provide prompt, student-friendly feedback. Two methods are used to model academic performance in dataset 1: (i) GPA prediction using regression and (ii) grade-category prediction using multi-class classification. We examine the performance of two powerful tree-based learners, Random Forest and Gradient Boosting, in a held-out test split. Gradient Boosting outperformed Random Forest (R2 0.924) in GPA regression with a R2 of 0.946 and RMSE of 0.212 GPA points. Gradient Boosting has the best macro F1 among tested baselines when it comes to grade categories, with the models achieving approximately 0.71 accuracy. Beyond model accuracy, the system focuses on usefulness: feature importance and correlation analysis are used to explain dominant drivers (e.g.Absences and study time are model signals that the web UI converts into short, actionable recommendations. We conclude by describing the complete implementation (React Vite frontend, FastAPI backend, REST endpoints) and discussing ethical considerations such as bias, privacy, and responsible communication of predictions.

DOI: http://doi.org/10.5281/zenodo.21455641

Explainable Heart Disease Prediction System Using Ma-chine Learning

Authors: Riya Kapil, Riyanshu Saini, Ashish Srivastava

Abstract: The majority of deaths worldwide are caused by heart disease. Doctors can use machine learning (ML) models to predict the risk of heart disease from routine exams and tests. The use of black-box ML models in healthcare is hindered by their lack of explanation for prediction. The proposed EHD-ML system combines effective ML models, such as gradient boosted trees and neural net-works, with techniques for explain ability that can be ap-plied to any model. SHAP, LIME, rule extraction, and counterfactual explanations are just a few of the things that are included. We cover dataset preparation, feature engineering, model training, interpretability pipelines, evaluation metrics like accuracy, AUC, F1, and calibra-tion, along with user-friendly explanations for clinicians, such as feature importance and patient-level explana-tions. We also outline the software and hardware design for deployment and suggest validation through retrospec-tive studies and prospective clinical trials. The key contri-butions include: (1) an end-to-end pipeline focused on explain ability for heart disease prediction, (2) a compar-ative analysis of interpretability methods and how accu-rately they reflect model predictions, and (3) user-centered explanation templates tailored for clinical use.

DOI: http://doi.org/10.5281/zenodo.21455793

Smart Farming Based On Ai-Based Crops Predictions

Authors: Hemlata, Jaishree Goyal, Manoj Pal

Abstract: Agriculture plays a crucial role in the economic development and food security of India. Predictable climate, soil variability, excessive fertilizer usage, and a lack of data-driven decision support systems often pose significant challenges for farmers in selecting suitable crops. The use of traditional farming practices is heavily dependent on experience and intuition, which may not always be in alignment with current environmental conditions and can result in low productivity and financial losses. This study presents a Smart Farming system that uses Machine Learning to accurately predict crops using key environmental and soil parameters such as Nitrogen (N), Phosphorus (P), Potassium (K), temperature, humidity, pH, and rainfall. A Random Forest classifier was trained using a dataset that is publicly available and contains 2200 samples and 22 crop classes. The proposed model's accuracy was approximately 96%, which demonstrates strong predictive performance for multi-class crop recommendations.

DOI: http://doi.org/10.5281/zenodo.21455972

A Modern Approach To Low-latency Peer-to-peer Communication Using Real-Time Video Conferencing

Authors: Manish Kumar, Paramjeet Singh, Hemlata

Abstract: Messaging is now part of everybody's life (1) in the world due to its ease and convenience for instant communication and simple to use. Chat facilities are available on almost all social networks worldwide. Existing chat applications have many common features. Here is a new chat application system called RTVC Application by the authors with new features such as videoconferencing and screen sharing. This RTVC application with meetings feature is a comprehensive communication tool designed for real- time messaging and audio/video conferencing. The application offers a modern and intuitive approach to communication, providing a seamless user experience. It is built using MongoDB (2), ExpressJS (3), ReactJS (4), and NodeJS (5), making it a powerful and flexible tool. The chat feature allows users to communicate via text messaging in real time, with the ability to create groups and add multiple users. The meeting feature enables users to host or join audio/video calls, with the option to mute/unmute audio or video and share their screen. The application's interface is clean and modern, providing a user-friendly experience. This RTVC chat application with meetings feature is ideal for remote teams, students, and individuals who require reliable and efficient communication tools. The application provides a versatile and flexible way to collaborate and stay connected, with its seamless user experience and intuitive design.

DOI: http://doi.org/10.5281/zenodo.21456058

Predictive Analysis Of Heart Disease Using Logistic Regression

Authors: Reshoo Devi, Shilpy Sharma, Raj Kumar

Abstract: Heart disease remains a major cause of death worldwide, accounting for a significant portion of annual mortality and posing significant challenges to healthcare systems. Preventing severe cardiac events requires timely and accurate diagnosis, but conventional diagnostic procedures are often time-consuming, costly, and dependent on expert clinical judgment, which may lead to variability in diagnosis. Intelligent systems that can assist clinicians in early disease prediction and risk assessment have been developed thanks to the increasing availability of medical data and advancements in machine learning (ML) and deep learning (DL). This research proposes a comprehensive Heart Disease Prediction System (HDPS) that utilizes machine learning and deep learning techniques to classify patients as either having heart disease or being healthy based on multiple clinical parameters. The system is developed using the UCI Cleveland Heart Disease dataset, from which 14 standard medical attributes were selected, including age, sex, chest pain type, cholesterol level, resting blood pressure, electrocardiographic results, and exercise-induced angina. The impact of data preprocessing strategies like feature selection, normalization, and outlier detection on predictive performance was analyzed using a multi-phase modeling approach. A sequential deep learning model was used along with multiple ML classifiers, including Logistic Regression, K-Nearest Neighbors, Support Vector Machine, Decision Tree, Random Forest, and XGBoost, for implementation and evaluation. Metrics used to assess model performance included accuracy, precision, recall, specificity, sensitivity, F1-score, and confusion matrix analysis. Proper preprocessing can significantly enhance model accuracy and stability, as demonstrated by experimental results. It has been confirmed by the results that predictive systems based on machine learning and deep learning can be reliable, efficient, and cost-effective clinical decision support tools for detecting heart disease early. Healthcare environments with limited resources can benefit from systems that can enhance diagnostic accuracy, reduce manual workload, and support timely medical intervention. The importance of data quality and preprocessing in medical analytics is highlighted in this study, which provides a foundation for future research that focuses on hybrid models, larger datasets, and real-time healthcare applications.

DOI: http://doi.org/10.5281/zenodo.21456108

CNN And RNN Are Predicting A High-frequency Bit Coin Trend

Authors: Vivek Kumar, Vinod Rathi, Vineet Salar

Abstract: Bit coin is a type of digital currency that is used for online transactions. It is a digital currency that does not exist in hard currency form. Our focus is on the distinction between a decentralized currency and a centralized currency, which means that all virtual currency users can acquire services without the aid of a third party. Due to their severe price volatility, the use of these crypto currencies has an impact on international relations and trade. A reliable method for estimating this price is urgently necessary due to the rapid variations in the prices of crypto currencies. The level of one main or central control over them has been significantly affected by price control by a number of organizations, affecting relationships with other businesses and international trade. In addition, the constant oscillations suggest that a more precise method of estimating this price is urgently required. Thus, using deep learning techniques such as the recurrent neural network (RNN) and the long short-term memory (LSTM), gated recurrent unit (GRU), which are effective learning models for training data, we must design a method for the accurate prediction of by considering various factors such as market cap, maximum supply and, volume, circulating supply. Python is used to write the proposed method and it is tested on benchmark datasets. It can be inferred from the results that the proposed method is capable of making reliable predictions. For the past ten years, academics in various fields have used neural networks as one of the intelligent data mining tools. The importance of stock market data cannot be overstated in today's economy. Forecasting methodologies can be divided into two types: linear (AR, MA, ARIMA, ARMA) and nonlinear models (ARCH, GARCH, Neural Network). To anticipate a company's stock price based on past prices, we employed Autoregressive Integrated Moving Average (ARIMA), Recurrent Neural Network, Long Short-Term Memory (LSTM), and Gated Recurrent Unit Deep Learning Architectures (GRU).

DOI: http://doi.org/10.5281/zenodo.21467365

Advanced Converter-Based Control for Power Quality Improvement in PV–Battery Grid Integration

Authors: Dilip Chauhan, Satyam Kumar Upadhyay, Sarvendra Kumar Singh

Abstract: This paper presents an analytical study on efficient power flow control in grid-connected photovoltaic and battery systems, emphasizing their capability to enhance power quality and ensure stable grid interaction. A comprehensive model is developed in MATLAB/Simulink, integrating PV generation with battery storage and shunt compensation to regulate voltage, mitigate harmonics and maintain optimal power exchange with the grid. The PV array is operated under variable irradiance conditions while the battery compensates for fluctuations through controlled charging and discharging. A robust maximum power point tracking (MPPT) algorithm ensures rapid convergence of the PV operating point enabling effective utilization of solar energy. The battery–converter interface is analyzed for voltage stability and current dynamics during abrupt load changes. Shunt inverters are investigated for reactive power support and harmonic suppression contributing to enhanced voltage regulation at the point of common coupling. Simulation results confirm that the proposed control framework achieves efficient power balancing among the PV array, battery and utility grid even under transient disturbances. Grid voltage and current waveforms remain well-synchronized and load-side power quality is preserved despite nonlinear demand. The study demonstrates that coordinated operation of PV and battery resources, supported by advanced control of interfacing converters provides a resilient and efficient solution for integrating renewable energy into low-voltage distribution networks. The findings offer practical insights for designing smart grid systems capable of sustaining reliable power delivery while maximizing renewable energy penetration.

Smart Healthcare Infection Prediction Using Machine Learning

Authors: Abhishek Kumar, Akshay Kumar, Rahul Kumar Patel, Abhishek Tyagi

Abstract: Smart healthcare infection prediction systems represent a transformative approach to improving hospital infection control by enabling the early detection and prediction of hospital-acquired infections (HAIs) among admitted patients. These systems leverage the computational power of artificial intelligence and machine learning models to analyze structured clinical data, including patient demographics, ICU admission status, duration of hospitalization, use of invasive devices such as ventilators and catheters, and underlying comorbidities. By identifying complex patterns and risk factors in real time, the system provides healthcare professionals with timely risk assessments that support proactive clinical decision-making and targeted intervention strategies. This research presents the design, implementation, and evaluation of a smart healthcare infection prediction system specifically developed for predicting HAIs within hospital environments. The study utilizes a dataset of patient records and applies machine learning models such as Gradient Boosting, Logistic Regression, and Random Forest to classify infection risk levels. The system is integrated with a web-based interface to facilitate ease of use for healthcare practitioners. Furthermore, the research examines key challenges including data quality, model interpretability, clinical reliability, and ethical considerations related to patient data privacy. The findings highlight the effectiveness of machine learning-driven approaches in enhancing patient safety, reducing infection rates, and optimizing hospital resource utilization, thereby contributing to more efficient and proactive healthcare management systems.

DOI: https://doi.org/10.5281/zenodo.21471420

Attacks & Preventing Measure On Cloud Robotics System, A Review with Cloud Heritage Technique

Authors: Ashutosh Kumar, Gaurav Singh Panwar, Gori Sharma, Jivesh Tiwari, Dr. Rajkumar

Abstract: This study investigates the vulnerability landscape of cloud robotics systems, identifying potential attack vectors that threaten the integrity, availability, and confidentially of robotic operations in a cloud environment. we delve into threats such as data branches, unauthorized access, intruders, and service disruptions within the cloud robotics paradigm. The research proposes a set of preventive measures to fortify this security technique, including the implementation of secure communication protocols, robust access controls, and an intrusion detection system. In this paper, we conduct a comprehensive study of security issues in the real-time cloud robotics environment. By analyzing different attack vectors in cloud robotics networks, we find attacks that manipulate the network resources, microarchitecture resources, and function parameters respectively.

DOI: https://doi.org/10.5281/zenodo.21471731

Fuzzy Logic Based Mathematical Models for Decision-Making Under Uncertainty

Authors: Professor Dr.T.Rama Rao, Assistant Professor M. Kalyani

Abstract: Making decisions amid uncertainties poses one of the central problems in such areas as financial risk analysis and innovative sustainable development of business organizations. Classical binary logical models are unable to properly reflect vagueness and imprecision typical of actual decision-making environment. This paper provides an in-depth study of mathematical fuzzy logic models used in decision making processes. In particular, we analyze the theoretical background of fuzzy set theory as well as its different modifications like intuitionistic, hesitant, and fuzzy N-bipolar soft sets and apply these models in the context of multi-criteria decision making. Our approach is based on the use of Mamdani-type fuzzy inference system in conjunction with genetic algorithm for fine-tuning of the parameters. Quantitative analysis conducted through simulation shows that the proposed fuzzy-genetic model possesses the classification accuracy of over 84%.

DOI: https://doi.org/10.5281/zenodo.21473972

Beyond the Instrument: A Human-Centric Forensic Framework for Questioned Document Authentication in the Digital Age

Authors: Assistant Professor Lakshya Bhardwaj, Assistant Professor Vaishali Vardiya, Ayan Khan, Raju Prajapati, Shivam Patel, Abushahma

Abstract: Powerful analytical technologies increasingly shape questioned document examination, yet the evidential meaning of a document still depends on human action, cognition, and behavior. Handwriting is a learned neuromuscular activity with identifiable regularities. Still, it is also variable across context, health status, emotional condition, and writing task, which makes source attribution an interpretive rather than purely instrumental problem (Hicklin et al., 2022; Singh et al., 2025; Nikolaychuk & Bila, 2023). At the same time, large validation work shows that forensic handwriting comparison can achieve useful accuracy, while still having nontrivial error and inconclusive rates and remaining vulnerable to cognitive bias, especially when judgments are made without structured safeguards (Hicklin et al., 2022; Cooper & Meterko, 2019; Kunkler & Roy, 2023). This paper proposes an equipment-independent, human-centric framework for questioned document authentication built on three linked pillars: the cognitive-scriptographic nexus, the material-semiotic interface, and the forensic intelligence paradigm. The first pillar treats handwriting as a behavioral biometric produced by distributed neural and motor processes rather than as a static pattern alone (Burgio et al., 2026; Qi et al., 2025; Ferrer-Ballester et al., 2017). The second pillar integrates line quality, pen lifts, spacing, pressure, and document substrate features into a single action-based interpretation model instead of separating physical and expressive evidence (Ellen et al., 2018; Cieśla, 2021; Kerniakevych-Tanasiichuk et al., 2021). The third pillar extends questioned document examination from case resolution to systematic intelligence generation through structured recording of forgery methods, alteration patterns, and cross-case linkages (Baechler & Margot, 2016; Baechler et al., 2012; Song et al., 2026). The literature supports such a framework most strongly when it is paired with explicit limits: contemporaneous standards for impaired writers, blind verification, contextual information management, and hybrid human-AI workflows rather than replacement of expert judgment (Singh et al., 2025; Bird & Yang, 2024; Can et al., 2026; El-Din, 2022). The main contribution is therefore not a rejection of instruments, but a structured interpretive model that makes human reasoning in questioned document examination more transparent, reproducible, and legally defensible.

Investigation of Lift Coefficient Characteristics for an Airfoil with High-Lift Devices Via Ansys Fluent

Authors: Van Huy Khuat, Le Thanh Nguyen, Van Viet Vo

Abstract: This paper presents the results of a survey on the change in the lift coefficient of the wing profile when using flaps on the Ansys Fluent application platform. The use of these flaps is extremely important. They not only ensure safe and efficient takeoff and landing but also enable rapid takeoff to approach and destroy targets even under runway damage. The paper investigates the change in the lift coefficient of the wing with flaps using Ansys Fluent software. Using this method, it is possible to determine the parameters corresponding to each type of flap, calculate the lift coefficient in each case, and identify the causes of these parameter changes.

DOI: https://doi.org/10.5281/zenodo.21488774

A Review on Sensor-Based Measurement Techniques in Modern Wearable and Intelligent Systems

Authors: Mehrdad Esmaeilipour

Abstract: Wearable and intelligent sensor systems have transformed modern healthcare, rehabilitation, and assistive technologies. Recent advances in soft sensors, biosensors, MEMS-based inertial units, optical and electrochemical sensing, and AI-driven signal processing have enabled accurate, real-time measurement of physiological, biochemical, and environmental parameters. This paper reviews the principles, measurement mechanisms, and applications of modern sensors—including electrochemical, optical, piezoelectric, inertial, and haptic sensors—based on recent literature. Furthermore, the paper highlights emerging trends such as intelligent soft sensors, multimodal biosensing, machine-learning–enhanced measurement, and smart wearable assistive devices for visually impaired users. The review concludes with future perspectives on sensor miniaturization, energy harvesting, and AI-driven autonomous measurement ecosystems.

DOI: https://doi.org/10.5281/zenodo.21493456

Village-Level Suitability Assessment For Litchi Cultivation In The Malwa Region Using Explainable Machine Learning And Geospatial Data

Authors: Dr. Pankaj Malik, Mishthi Patodia, Vedant soni, Deepika Kumari, Pragati Agrawal, Jaiswal Tanmay

Abstract: Litchi is a high-value fruit crop traditionally cultivated in regions with favorable climatic and soil conditions. Expanding litchi cultivation into non-traditional areas such as the Malwa region of Madhya Pradesh requires accurate identification of suitable locations to minimize cultivation risks and maximize productivity. This study proposes a village-level suitability assessment framework that integrates geospatial data, climatic variables, soil characteristics, groundwater availability, and satellite-derived vegetation indices with Explainable Machine Learning (XML) techniques. Environmental and agricultural data were collected from multiple sources, including Sentinel-2 imagery, Soil Health Card records, meteorological datasets, and groundwater databases. Several machine learning algorithms, namely Random Forest, XGBoost, LightGBM, CatBoost, and Support Vector Machine, were trained to predict the suitability of villages for litchi cultivation. To enhance transparency and interpretability, SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-Agnostic Explanations) were employed to identify the key factors influencing suitability predictions. The experimental results demonstrated that the XGBoost model achieved the highest classification performance with an accuracy of 94.2%, precision of 93.6%, recall of 92.8%, F1-score of 93.2%, and ROC-AUC of 0.96, outperforming the other evaluated models. SHAP analysis revealed that winter temperature, annual rainfall, soil organic carbon, groundwater depth, and irrigation availability were the most influential parameters affecting litchi suitability. The generated village-level suitability maps identified several highly suitable zones within the Indore, Dewas, and Ujjain districts, while areas experiencing higher temperatures and limited water resources were classified as marginally suitable or unsuitable. The proposed framework provides a transparent and data-driven decision support system for farmers, horticulture planners, and policymakers, facilitating climate-resilient expansion of litchi cultivation in the Malwa region.

DOI: http://doi.org/10.5281/zenodo.21504453

Artificial Intelligence and Machine Learning in Banking Sector-An Overview

Authors: Dr. Vaishali Doshi

Abstract: The banking sector has been transformed from people centric to customer centric. By adopting a holistic approach bank are trying to meet customer demands and expectations. Here the Artificial intelligence and machine learning technologies are acting as savior. They are useful in processing large volume of data and anticipate the current market trends. These technologies are also useful for credit management, understanding behavior and pattern of customers, validating customers creditworthiness thus reducing the chances of any bank default, marketing, risk-management, process automation, providing information on banks solvency, trading activities, regulatory compliance etc. Therefore, these technologies are playing a very crucial role in banking sector. The paper analyzes role of AI and ML in banking sector by focusing on their advantages as well as challenges in implementing these technologies.

Smart Grid Power Enhancement Using Pv-Wind Hybrid Systems With Energy Storage

Authors: Satyam Kumar Upadhyay, Shamsher kumar Bharati

Abstract: This study presents a comprehensive approach to enhance power quality and efficiency in a grid-connected photovoltaic (PV) and wind energy (WE) hybrid system integrated with energy storage systems (ESS) and electric vehicles (EVs). By leveraging advanced Adaptive Neuro-Fuzzy Inference System (ANFIS) based Maximum Power Point Tracking (MPPT) techniques, the system dynamically maximizes the power output from PV arrays under varying environmental conditions, thereby optimizing energy utilization. The integration of ESS stabilizes the power flow, compensates for renewable intermittency, and supports grid reliability. Additionally, an electric vehicle aggregator (EVA) framework is designed to enable bidirectional power flow—vehicle-to-grid (V2G) and grid-to-vehicle (G2V)—facilitating uninterruptible power supply and efficient load demand management. Various power quality enhancement techniques, including advanced controllers and power electronic devices, are incorporated to mitigate voltage fluctuations, harmonics, and other disturbances inherent in renewable energy integration. Simulation results demonstrate significant improvements in voltage stability, reduced power losses, and harmonic distortion mitigation. The proposed integrated system offers a robust solution for sustainable, efficient, and reliable renewable energy integration into the grid, advancing the role of EVs as active participants in energy management and promoting optimal utilization of renewable resources through ANFIS MPPT-based control strategies.

Bidirectional Power Flow Optimization In Vehicle-to-Grid (V2G) Enabled Electric Vehicles

Authors: Adarsh Kumar Bhardwaj, Jaya shukla

Abstract: The increasing integration of electric vehicles (EVs) into smart grids necessitates advanced control strategies for efficient bidirectional power flow management. Vehicle-to-Grid (V2G) systems enable EVs to act as distributed energy resources, supporting grid stability and economic operation. However, challenges such as stochastic load variations, renewable intermittency, and battery degradation limit the effectiveness of conventional control approaches. This paper proposes a multi-objective stochastic model predictive control (MO-SMPC) framework integrated with adaptive power sharing between battery and supercapacitor storage systems for optimal V2G operation. The proposed method simultaneously minimizes electricity cost, grid power fluctuations, and battery aging while ensuring system constraints. A novel adaptive droop-based predictive control strategy is introduced to dynamically allocate power between battery and supercapacitor, enhancing transient response and extending battery life. Stochastic modeling is incorporated to address uncertainties in load demand, EV availability, and solar generation. MATLAB-based simulations under multiple operating scenarios demonstrate improved peak shaving, reduced operational cost, and enhanced battery lifespan compared to conventional battery-only systems. Results show up to 40% reduction in cost, 30% improvement in grid stability, and significant mitigation of battery stress. The proposed framework offers a scalable and practical solution for intelligent energy management in next-generation smart grids.

Lemongrass (Cymbopogon Citratus) As A Phytogenic Feed Additive For Improving Growth Performance, Gut Health And Immunity In Poultry: A Review

Authors: P. Radhika

Abstract: The restriction on antibiotic growth promoters in poultry production has increased interest in natural feed additives that can enhance productivity while maintaining animal health. Lemongrass (Cymbopogon citratus) is a promising phytogenic additive due to its rich content of bioactive compounds, including citral, geraniol, flavonoids, phenolic acids and terpenoids. These compounds possess antimicrobial, antioxidant, anti-inflammatory and immunomodulatory properties that may improve growth performance, intestinal health and disease resistance in poultry. This review summarizes the effects of lemongrass supplementation on production efficiency, gastrointestinal health, antioxidant status and immune responses in broiler chickens and other poultry species. Available evidence indicates that moderate dietary inclusion of lemongrass or its essential oil improves body weight gain, feed conversion efficiency, antioxidant capacity and immune function, although responses vary depending on dosage, formulation and experimental conditions. Further research is required to standardize supplementation levels and clarify molecular mechanisms. Overall, lemongrass represents a potential natural alternative to antibiotic growth promoters in sustainable poultry production.

Arrival Rate Uncertainty and the Solution Method in the Queueing Model Analysis

Authors: Pushpandra Kumar, Parul Agarwal

Abstract: While queueing systems with random components have been thoroughly examined, this work presents an uncertain queueing model that takes the degree of customer behavior belief into account. Understanding the dynamic behavior of the underlying processes is the primary objective of mathematical modeling and analysis of queueing systems, allowing for intelligent and well-informed management choices. This study contributes to the body of research on rational consumers' decision-making in an uncertain setting where certain system parameters are random variables whose actual value is unknown to the consumers at the time of decision-making. In particular, models with a random arrival rate are of interest. Customers arrive in groups in this approach, and service and inter-arrival times are handled as unknown variables. The busy index and busy time are also explored, and the analytical formulae and uncertainty distribution are produced, respectively, in order to assess the performance of this suggested uncertain queueing model. Furthermore, several numerical examples are provided to clarify the results that were reached.

DOI: https://doi.org/10.5281/zenodo.21533080

Technology As The Fourth Pillar A Conceptual Framework For ESGT Governance In The Age Of Artificial Intelligence

Authors: Rudy Shoushany

Abstract: Environmental, Social and Governance (ESG) analysis has become the dominant grammar of corporate non-financial accountability, yet its tripartite architecture was conceived in 2004, before the smartphone, large scale cloud computing, platform data economies and generative artificial intelligence reshaped organisational risk. This paper advances a governance first theory of ESGT, in which technology is elevated to an explicit fourth pillar, denoted T, encompassing artificial intelligence governance, algorithmic accountability, data governance, digital ethics and responsible technology, and cybersecurity. The framework is organised around an auditable pillar principle, namely that a governance dimension earns recognition only if it can be measured and assured, and it therefore anchors the T pillar in a mature stack of standards, including ISO/IEC 42001, the European Union Artificial Intelligence Act, the NIST Artificial Intelligence Risk Management Framework, COBIT 2019 and ISO/IEC 27001. Particular weight is placed on COBIT 2019, whose explicit separation of governance from management, seven governance components and design factor tailoring supply the architectural template through which the T pillar is structured and made auditable. The contribution is threefold. First, the paper documents the technology gap across the major reporting standards, drawing on recent evidence that fewer than one in ten large European firms disclose artificial intelligence as a risk category. Second, it answers the central objection that technology governance is unmeasurable by mapping the T pillar to auditable instruments and by drawing on the scholarly construct of Corporate Digital Responsibility (Lobschat et al., 2021) for its normative content. Third, it engages the strongest counterarguments, namely dilution, framework proliferation, redundancy and rating divergence, and specifies the empirical conditions under which a separate pillar is justified rather than mere integration. The framing is global and standard agnostic, intended to inform standard setters, boards and investors.

DOI: http://doi.org/10.5281/zenodo.21533094

The Leap Of Faith And The Sacrifice Of The Self: A Kierkegaardian Reading Of Lars Von Trier’s Breaking The Waves

Authors: Panagiota Boumpouli

Abstract: Lars von Trier's Breaking the Waves (1996) has generated extensive scholarly debate regarding its treatment of religion, gender, sexuality, sacrifice, and transcendence. While previous studies have primarily interpreted the protagonist, Bess McNeill, either as a Christ-like figure or as a victim of patriarchal religious ideology, this article proposes an alternative reading grounded in Søren Kierkegaard's philosophy of existence. It argues that Bess embodies the Kierkegaardian knight of faith, whose radical commitment to divine love transcends universal ethical norms through the paradoxical movement of faith. Drawing upon Fear and Trembling, Philosophical Fragments, and Stages on Life's Way, the article examines how Bess's existential journey unfolds through Kierkegaard's three stages of existence—the aesthetic, the ethical, and the religious. Particular attention is devoted to the concepts of the leap of faith, infinite resignation, paradox, and agapic love, demonstrating that Bess's sacrifice cannot be adequately understood as passive submission or psychological pathology. Instead, her actions constitute a conscious existential choice motivated by unconditional love and an absolute relation to God. Furthermore, the article explores von Trier's symbolic use of sexuality, corporeality, suffering, and the recurring motif of the church bells as theological signs that challenge institutional religion while affirming divine grace. Ultimately, Breaking the Waves presents transcendence not as an escape from human existence but as its deepest realization through suffering, love, and faith. By bringing Kierkegaard's existential theology into dialogue with contemporary film analysis, this study offers a new interpretation of von Trier's film, positioning Bess not merely as a tragic heroine but as an authentic religious subject whose paradoxical faith transforms both herself and those around her.

Optimal Decision-Making Under Model Misspecification: A Unified Framework for Robust and Learning-Based Optimization

Authors: Vaivaw Kumar Singh

Abstract: Decision making under uncertainty is one of the main difficulties not only in economics, finance, operations research, engineering, and artificial intelligence but also pretty much anywhere. Most of the traditional optimization techniques work on the assumption that the underlying models are a perfect reflection of the real-world systems, but, the reality is that model misspecification is an everyday event due to incomplete information, structural changes, and uncertain environments (Hansen & Sargent, 2008). These types of errors might dramatically lower the quality of decisions and the performance of the system. This paper starts from the existing optimization methods' problems and derives a joint structure that combines robust optimization and learning-based optimization for handling model misspecification. Besides reviewing the literature that supports the four types of decision-making methods, i.e. robust optimization, reinforcement learning, adaptive decision-making, and distributionally robust optimization, the study also presents a balanced performance analysis of these methods at various stages of development and takes into account different modes of fuzziness and uncertainty. The results show that robust optimization can effectively deal with uncertainty and result in conservative decisions; in contrast, learning-based methods improve adaptability but are exposed to shifts in the distributions and the presence of structural model errors (Sutton & Barto, 2018). Our setup implements both awareness of uncertainty and the ability to learn new things in decision making process and results in solutions that have adequate levels of both robustness and flexibility. Our research offers to the body of optimization knowledge a system upon which conceptual support can be built for reliable and adaptive decision making in a constantly changing environment. The structure may find its facets in the areas of finance healthcare supply chain management, autonomous systems, and artificial intelligence.

Explainable Medical Diagnosis System Using Machine Learning and Explainable AI

Authors: Navya Shree K V, Sagar D, H S Shashank, Sinchana S Y, Professor Dr. Dilshad Begum, Professor Dr. T Venkatesh

Abstract: Artificial Intelligence (AI) has significantly improved healthcare by enabling accurate and timely disease diagnosis; however, the limited interpretability of deep learning models restricts their adoption in clinical practice. This paper proposes an Explainable Medical Diagnosis System (EMDS) that integrates machine learning, deep learning, and Explainable Artificial Intelligence (XAI) to deliver accurate, reliable, and interpretable disease prediction. The proposed framework consists of six modules: medical data acquisition, data preprocessing, feature engineering and selection, hybrid disease classification, explainable AI analysis, and clinical decision support. During preprocessing, missing values are imputed, medical images are enhanced using Contrast Limited Adaptive Histogram Equalization (CLAHE), and class imbalance is addressed using the Synthetic Minority Oversampling Technique (SMOTE). Feature optimization is performed through Recursive Feature Elimination (RFE) and Mutual Information. The hybrid classification model combines EfficientNetV2 for deep image feature extraction, TabNet for structured electronic health record learning, and XGBoost for disease classification, while a stacking ensemble improves predictive accuracy and generalization. Model interpretability is achieved using SHapley Additive exPlanations (SHAP), Local Interpretable Model-Agnostic Explanations (LIME), and Grad-CAM++, providing both global feature importance and patient-specific visual explanations. Experimental evaluation using benchmark medical datasets and five-fold cross-validation demonstrates superior performance over conventional approaches in terms of accuracy, precision, recall, F1-score, and AUC. The proposed EMDS provides a scalable, interpretable, and trustworthy framework for AI-assisted clinical decision-making in modern healthcare systems.

Synthesis, Spectral Characterization, and Biological Evaluation of 2-((2-(7H-pyrrolo[2,3-d]pyrimidin-4-yl)hydrazono)methyl)-5-fluorophenol and Its Transition Metal Complexes

Authors: Hanumant Rananaware, M. A. Badgujar

Abstract: A new hydrazone ligand, 2-((2-(7H-pyrrolo[2,3-d]pyrimidin-4-yl)hydrazono)methyl)-5-fluorophenol, was synthesized via the condensation reaction of 5-fluorosalicyldehyde and 7H-pyrrolo[2,3-d]pyrimidine-4-carbaldehyde hydrazone. The ligand was characterized using various spectroscopic techniques, including FTIR, NMR (¹H, ¹³C), UV-Vis, and mass spectrometry, confirming its molecular structure. The ligand was then complexed with several transition metals (Fe²⁺, Cu²⁺, Ni²⁺, Co²⁺) to form metal complexes. The resulting metal complexes were characterized by elemental analysis, FTIR, UV-Vis spectroscopy, and magnetic susceptibility studies. The results indicated that the metal ions coordinated with the ligand via the nitrogen atom of the hydrazone group and the oxygen atom of the phenolic group, forming stable 1:2 complexes with octahedral or square planar geometries, depending on the metal ion. Biological studies of the ligand and its metal complexes demonstrated significant antimicrobial activity against both Gram-positive and Gram-negative bacteria, with the metal complexes showing superior activity compared to the free ligand. In addition, the complexes exhibited notable antioxidant properties, with the Cu²⁺ and Co²⁺ complexes showing the highest scavenging activity. The enhancement of biological activity upon metal coordination suggests that the metal complexes may have potential therapeutic applications as antimicrobial and antioxidant agents. These findings highlight the importance of metal coordination in modulating the biological properties of hydrazone-based ligands, providing a foundation for future studies in medicinal chemistry and bioinorganic chemistry.

DOI: https://doi.org/10.5281/zenodo.21552783

Synthesis, Spectroscopic Characterization and Antibacterial Activities of 4-Chlorobenzylidene)-2-(2-(hydroxyimino)-1,2-diphenylethylidene)hydrazine-1-carbothioamide and its Metal Complexes

Authors: Sandip Thube, M. A. Badgujar

Abstract: In this study, we report the synthesis, detailed spectroscopic characterization, and antibacterial evaluation of the Schiff base ligand, 4-(Chlorobenzylidene)-2-(2-(hydroxyimino)-1,2-diphenylethylidene)hydrazine-1-carbothioamide, and its coordination complexes with selected transition metals. The ligand was synthesized via a condensation reaction and was further complexed with metal ions, including Cu(II), Ni(II), and Zn(II), under controlled conditions. Structural elucidation of both the ligand and its metal complexes was achieved using Fourier-transform infrared (FT-IR), nuclear magnetic resonance (NMR), ultraviolet-visible (UV-Vis), and mass spectroscopic techniques, with supplementary single-crystal X-ray diffraction performed for select complexes. The spectral data confirmed successful complexation, with notable shifts in functional group peaks indicating coordination between the ligand and metal ions. The antibacterial activity of the synthesized compounds was evaluated against Gram-positive and Gram-negative bacterial strains, revealing enhanced activity for the metal complexes compared to the free ligand, particularly against Escherichia coli and Staphylococcus aureus. The findings suggest that metal coordination may amplify the biological efficacy of the ligand, offering the potential for future antimicrobial drug development.

DOI: https://doi.org/10.5281/zenodo.21553433

Multi-Class Litchi Fruit Disease Detection Using a Novel Hybrid Deep Learning and Image Processing Framework with Explainable Artificial Intelligence

Authors: Dr. Pankaj Malik, Rahul Verma, Ashish Chourey, Sameer Tandon, Riddhi Yadav

Abstract: Litchi (Litchi chinensis Sonn.) is a high-value tropical fruit crop that is highly susceptible to various diseases, including anthracnose, fruit rot, pericarp browning, sooty mold, and algal spot, which significantly reduce fruit quality, market value, and yield. Accurate and timely disease diagnosis is essential for effective disease management and sustainable litchi production. Traditional visual inspection methods are labor-intensive, subjective, and often incapable of detecting diseases at early stages. Recent advances in deep learning have demonstrated promising results for automated plant disease detection; however, conventional Convolutional Neural Networks (CNNs) primarily focus on local features and often fail to capture global contextual information. Furthermore, the lack of model interpretability limits their practical adoption in smart agriculture applications. To address these challenges, this study proposes a novel Hybrid CNN–Vision Transformer (HCVT) framework for multi-class litchi fruit disease detection and severity assessment. The proposed model integrates EfficientNet-B3 and Vision Transformer (ViT) architectures through an attention-based feature fusion mechanism, enabling simultaneous extraction of local lesion characteristics and global contextual representations. The framework incorporates comprehensive image preprocessing, data augmentation, explainable artificial intelligence (XAI) techniques using Grad-CAM and SHAP, and a disease severity estimation module based on lesion segmentation. Experimental evaluation was conducted on a multi-class litchi disease dataset comprising healthy and diseased fruit images. The proposed HCVT model achieved an overall classification accuracy of 97.63%, precision of 97.31%, recall of 97.05%, F1-score of 97.18%, and AUC of 0.983, outperforming standalone CNN, Vision Transformer, and conventional hybrid models. Five-fold cross-validation confirmed the robustness and generalization capability of the framework. Grad-CAM visualizations successfully localized disease-affected regions, while SHAP analysis identified lesion texture, color variation, and infected area characteristics as the most influential features. Additionally, the proposed severity estimation module effectively quantified disease progression using lesion-area analysis. The results demonstrate that the proposed HCVT framework provides a highly accurate, interpretable, and reliable solution for automated litchi disease diagnosis. The integration of deep learning, transformer-based representation learning, explainable AI, and severity estimation offers significant potential for deployment in smart agriculture, precision farming, and future UAV-based crop health monitoring systems.

DOI: https://doi.org/10.5281/zenodo.21618962

Response Surface Methodology and Particle Swarm Optimization of Biodiesel Production from Waste Chicken Fat Oil Using KOH-Activated Chicken Bone Catalyst.

Authors: Nnali-Uroh Emmanue, Oluikpe Victor Amadi, Kolawole Idowu Sunday, Akpa Umahi Nwaeze

Abstract: This study investigates the optimization of biodiesel production from waste chicken fat oil using a KOH-activated chicken bone catalyst, employing an integrated experimental–computational approach. Response Surface Methodology (RSM), combined with Particle Swarm Optimization (PSO), was used to optimize five key process variables: catalyst dosage (0.5–2.5 wt%), reaction temperature (30–70 °C), reaction time (1–5 hours), methanol-to-oil ratio (6:1–14:1), and agitation speed (100–500 rpm). The optimal conditions predicted by RSM and PSO were: catalyst dosage of 1.30 wt%, methanol-to-oil ratio of 13.56:1, reaction time of 3.3 hours, and temperature of 77 °C. The resulting biodiesel yield was 97.42%, with a kinematic viscosity of 5.978 mm² s⁻¹ and a cetane index of 63.27, which meet the acceptable limits set by ASTM D6751 and EN 14214 standards. The study focused on the interactions of process variables, where the highest biodiesel yield was achieved by simultaneously increasing catalyst dosage and methanol ratio, with moderate temperature and extended reaction times. Statistical analysis using ANOVA revealed that the interaction model outperformed the linear model, with R² values of 0.6168 for biodiesel yield, 0.5233 for kinematic viscosity, and 0.5878 for cetane index, confirming the significance of interaction effects in the biodiesel production process. This research demonstrates that waste chicken fat oil, a low-cost feedstock, can be effectively utilized for biodiesel production. The optimized process conditions and integration of PSO for multi-response optimization provide a scalable and sustainable solution for biodiesel production from poultry-derived waste oils, contributing to waste-to-energy conversion

DOI: https://doi.org/10.5281/zenodo.21620524

Evaluating Trends in Agricultural Budgeting: A Five-Year Review Against Current Projections

Authors: Assistant Professor Dr.N.L.Srikanth, Assistant Professor Mr.M.Sravansai, Professor Dr.B.Sivakumar, Assistant Professor Mr.P.Sivanagesh, Assistant Professor Ms.M.Bhavani, Assistant Professor Dr.N.Madhu

Abstract: Agricultural budgeting is pivotal to advancing both the growth and long-term viability of the agriculture sector. This research presents a comparative evaluation of the projected agricultural budget for 2025–26 to the actual allocations made over the previous five fiscal years (2020–21 to 2024–25). Through a detailed assessment of trends in both recurring and capital expenditures, the study reveals notable shifts in financial prioritization across sub-sectors such as crop production, irrigation systems, agricultural research, and rural credit programs. Data were compiled from government budget records, financial audits, and official expenditure statements issued by the Ministry of Agriculture. Although overall funding for agriculture has shown a nominal upward trend annually, critical components like agricultural extension services and climate-resilient infrastructure have experienced irregular investment patterns. The study also identifies an average gap of 12.5% between projected and actual spending, with budget overestimations more frequently occurring during periods marked by political change or external economic disturbances. Furthermore, the research highlights imbalances between central and state-level contributions, pointing to inconsistencies in the distribution of resources. These findings emphasize the necessity for more reliable budget forecasting tools, improved coordination among implementing bodies, and flexible budgeting practices that can respond effectively to evolving sectoral needs. The study concludes by offering policy recommendations aimed at reinforcing fiscal responsibility, promoting transparency in the use of public funds, and aligning agricultural budgets more strategically with development goals and climate adaptation priorities.

DOI: https://doi.org/10.5281/zenodo.21638111

Isolation and Spectroscopic Characterization of Alkaloids from the Methanolic Leaf extract of Adhatoda vasica Nees

Authors: Ruchi Upadhya, Varun Jain

Abstract: Background: Adhatoda vasica Nees (Acanthaceae), commonly known as Vasaka, is an important medicinal plant extensively used in traditional Ayurvedic and Unani systems for the treatment of respiratory disorders. The therapeutic efficacy of the plant is largely attributed to the presence of quinazoline alkaloids, particularly vasicine and vasicinone. Objective: The present investigation aimed to isolate and characterize alkaloidal constituents from the methanolic leaf extract of A. vasica using chromatographic and spectroscopic techniques. Methods: Dried leaf powder was extracted with methanol by maceration, followed by acid–base fractionation to obtain alkaloid-enriched fractions. Preliminary screening was performed using thin-layer chromatography (TLC) and Dragendorff’s reagent. Isolation of alkaloids was achieved through silica gel column chromatography and preparative TLC. Structural characterization of the isolated compounds was carried out using UV–Visible spectroscopy, FT-IR, mass spectrometry (MS), and NMR analyses. Results: Methanolic extraction yielded an alkaloid-rich fraction that exhibited positive Dragendorff’s test. TLC profiling revealed multiple alkaloidal constituents with distinct Rf values. Chromatographic purification afforded a major alkaloidal fraction (SA-2), obtained as a pale-yellow amorphous powder with a melting point of 311–315°C. High-resolution mass spectrometry displayed a molecular ion peak at m/z 270.24. FT-IR analysis indicated characteristic absorption bands corresponding to hydroxyl, aromatic C–H, carbonyl, and olefinic functionalities. UV spectra showed absorption maxima at 280 and 443 nm. Spectral data were consistent with quinazoline alkaloid derivatives reported from A. vasica. Conclusion: The developed extraction and purification strategy effectively isolated alkaloidal constituents from A. vasica leaves. Spectroscopic evidence confirmed the presence of bioactive quinazoline alkaloids, supporting the medicinal significance of the plant and its potential for pharmaceutical applications.

Privacy-Preserving Deep Learning Framework For Secure Predictive Analytics In Industrial IIoT

Authors: Ms. Neha Yadav, Dr. Nitin Kumar

Abstract: The rapid adoption of the Industrial Internet of Things (IIoT) will continue to transform industrial operations by enabling real-time monitoring, intelligent automation, and data-driven decision-making. However, the increasing deployment of interconnected sensors, controllers, and smart devices will expose industrial systems to significant cyber security threats, data breaches, and privacy concerns. Conventional predictive analytics approaches will often fail to provide robust security while maintaining high prediction accuracy in dynamic industrial environments. Therefore, this study will propose a Secure Deep Learning Framework for Predictive Analytics in Industrial Internet of Things (IIoT) that will integrate advanced deep learning models with multi-layer security mechanisms to enhance both predictive performance and data protection.

DOI: http://doi.org/10.5281/zenodo.21641494

Simulation Of The AL-31F Engine Fuel System Using An Electrically Driven Fuel Pump In MATLAB/Simulink

Authors: Vu Van Dong, Dong Tat Dat, Truong Huu Noi

Abstract: This paper presents a simulation study of the АL-31F engine fuel system using MATLAB/Simulink with the Simscape/SimHydraulics library, focusing on the main fuel supply branch, including the centrifugal pump, the gear pump, and the potential application of an electrically driven fuel pump. The simulation results show that the centrifugal pump exhibits a noticeable variation in flow rate under throttling and changes in rotational speed, whereas the gear pump maintains an almost stable flow rate, is less dependent on outlet pressure, and is not suitable for regulation by throttling alone. On this basis, the use of an electric drive for the fuel pump, especially for the gear pump, is considered a promising approach. It enables fuel flow regulation through pump rotational speed, thereby reducing fuel recirculation, limiting hydraulic losses, and improving the controllability of the fuel system.

DOI: http://doi.org/10.5281/zenodo.21642223

A Queueing System with Servers Disguised as Customers

Authors: Parul Agarwal, Pushpandra Kumar

Abstract: Customer service has been modeled and optimized in many real-world scenarios using queuing theory. In this study, we suggest a new model that is inspired by the occurrence of pseudo progression in cancer, where the length of a line appears to expand for a while before decreasing more quickly. We consider servers to be "disguised" as customers when they join the line. We use matrix analytic techniques to determine the general equations for this model and use numerical simulations to show how it functions.

DOI: https://doi.org/10.5281/zenodo.21643248

Carbon Nanotube-Reinforced Aluminum A357 Nanocomposite by Using Stir-Casting

Authors: Rutuja Sunil Jadhav, Professor Dr. Pankaj P. Awate

Abstract: Aluminum matrix nanocomposites have attracted considerable attention in recent years due to their superior mechanical properties, lightweight nature, and potential for advanced engineering applications. Among various aluminum alloys, A357 is widely used in the automotive, aerospace, and marine industries because of its excellent castability, good corrosion resistance, and favorable strength-to-weight ratio. However, the mechanical performance of unreinforced A357 alloy can be further enhanced through the incorporation of nanoscale reinforcements such as multi-walled carbon nanotubes (MWCNTs). In the present investigation, A357 aluminum alloy reinforced with different weight fractions of MWCNTs was successfully fabricated using the stir casting technique. Stir casting was selected owing to its simplicity, low production cost, and suitability for large-scale manufacturing. The fabricated nanocomposites were evaluated to determine the influence of MWCNT reinforcement on their mechanical properties, including ultimate tensile strength, yield strength, hardness, and ductility. In addition, microstructural characterization was carried out using optical microscopy and scanning electron microscopy (SEM) to examine the dispersion of MWCNTs, grain refinement, and the quality of interfacial bonding between the matrix and reinforcement. The experimental results revealed that the incorporation of MWCNTs significantly enhanced the mechanical performance of the A357 alloy up to an optimum reinforcement level. Maximum improvement was observed at approximately 1.0 wt.% MWCNT, owing to the uniform distribution of nanotubes, effective load transfer, and grain refinement. Further addition of MWCNTs resulted in a slight reduction in mechanical properties because of nanotube agglomeration and increased porosity. The developed A357–MWCNT nanocomposites exhibit improved strength and hardness while maintaining acceptable ductility, making them promising candidates for lightweight structural components in automotive, aerospace, and other high-performance engineering applications.

DOI: https://doi.org/10.5281/zenodo.21643609

Prioritizing Risk Factors in Bridge Construction Projects: Evidence from the Gwalior-Chambal Region, India

Authors: Research Scholar Gauri Verma, Professor Dr. Manoj Kumar Trivedi

Abstract: Bridge construction projects combine structural complexity, long delivery periods, multi-party coordination, and exposure to uncertain geotechnical and hydrological conditions. This study identifies and prioritizes the risks affecting bridge projects in the Gwalior-Chambal region of Madhya Pradesh, India. A cross-sectional questionnaire measured 30 risk factors across technical, financial, construction, environmental, managerial, and legal/regulatory categories. Valid responses from 150 contractors, consultants, government engineers, project-management consultants, and other professionals were analyzed using descriptive statistics, the Relative Importance Index (RII), Cronbach's alpha, Pearson correlations, and one-way ANOVA. The instrument demonstrated excellent internal consistency (alpha = .915). Technical risk ranked first (mean = 3.7773; RII = .7555), followed by environmental and construction risk. Design errors and changes were the highest individual risk (RII = .8427), followed by inappropriate construction methods (RII = .8253); delayed payments, poor site management, and flooding each recorded an RII of .8080. Construction and environmental risks were strongly correlated (r = .769), as were financial and legal/regulatory risks (r = .768). Risk effects were greatest for time (RII = .7947) and quality (RII = .7846). Perceptions were stable across organization, designation, experience, and project type, but differed partly by the number of completed projects. The resulting framework links regional evidence, statistical priority, assigned responses, and continuous monitoring, offering a practical basis for more resilient bridge delivery.

DOI: https://doi.org/10.5281/zenodo.21644013

Impact of Project Smart on the Reading Proficiency of San Guillermo Elementary Learners Through Data-Driven Assessment and Strategic Instruction

Authors: Jonna May D. Casalme, Argel R. Doctora, Rocelyn D. Tenorio

Abstract: This action research study examined the impact of Project SMART on learners' reading proficiency at San Guillermo Elementary School during the 2024–2025 school year. This research utilizes a descriptive-quantitative approach. Moreover, the study focused on the 107 learners from Grades 1–6 who were assessed through pre- and post-test reading tests in English and Filipino. The SMART project intervention focuses more on data-driven assessment, meaning teachers use data to guide their instruction and help students become proficient readers. Additionally, they employ strategic interventions to help learners fully develop their decoding and reading proficiency. This phase enabled teachers and learners to tailor reading to individual learners' needs. Pre-test results revealed that a large proportion of pupils are slow, struggling, and non-readers, while 23.26% are fast readers in English and 28.04% in Filipino. Following the implementation of the guided oral reading approach, reading drills were conducted. Post-test data revealed a notable increase in fast readers, 36.64% in English and 37.38% in Filipino. There is also a significant decrease in non-reader learners in both subjects; however, there is still one. In line with the Department of Education’s thrust toward foundational literacy, the study concluded that Project SMART serves as an impactful, sustainable framework for developing progressive and independent readers. It is recommended to monitor progress further, modify instruction, and strengthen community partnerships to sustain reading growth and institutionalize the gains of the SMART achievement project.

DOI: http://doi.org/10.5281/zenodo.21671493

Mechanisms, Plant Species and Prospects of Phytoremediation in Managing Heavy Metal Contaminated Soils

Authors: Ogunsumi Akintunde Israel, Awodiran Festus Tunde

Abstract: Heavy metal pollution of soils in Nigeria, driven by activities such as mining, industrial discharge, waste disposal, and agricultural practices, poses a severe threat to environmental health, food safety, and human well-being. Conventional remediation methods are often costly, invasive, and environmentally disruptive. Phytoremediation, a green and cost-effective technology using plants to extract, stabilize, or degrade contaminants has emerged as a promising alternative for addressing soil contamination in Nigeria. This review explores the mechanisms of phytoremediation, including phytoextraction, phytostabilization, phytovolatilization, and rhizofiltration, with a focus on their application to heavy metals such as lead (Pb), cadmium (Cd), arsenic (As), and mercury (Hg). Several plant species indigenous or adaptable to Nigerian soils, including Chromolaena odorata, Chromolaena odorata, Brassica juncea, Helianthus annuus, and Tithonia diversifolia have demonstrated phytoremediation potential. Despite promising research, major knowledge gaps remain in understanding plant-metal interactions under local soil and climatic conditions, the long-term sustainability of remediation outcomes, and strategies for integrating phytoremediation into national environmental policies. Challenges such as low biomass yield, metal toxicity to plants, and limited public awareness hinder large-scale implementation. Future research should focus on genetic improvement of hyperaccumulators, synergistic use of soil amendments and microbes, and development of region-specific phytoremediation models. Strengthening collaboration among researchers, government agencies, and local communities is crucial to fully harness the potential of phytoremediation for sustainable soil management and environmental restoration in Nigeria.

DOI: https://doi.org/10.5281/zenodo.21770493

Comparative Analysis of Password-based, Multi-factor, and Biometric Authentication Methods and their Effectiveness in Reducing Unauthorized Access

Authors: Durojaye Emmanuel Olatunji, Akinola Daniel Adeoye, Ajayi Titus Ileriayo, Adebayo Suliat Precious, Adewale Joseph Adebiyi

Abstract: Software security has become increasingly important as organizations and individuals continue to depend on digital systems for communication, financial transactions, healthcare, education, and other critical services. One of the primary mechanisms used to protect these systems from unauthorized access is user authentication. Traditional password-based authentication remains the most widely adopted authentication method because of its simplicity and low implementation cost. However, the growing sophistication of cyberattacks, including phishing, brute-force attacks, credential stuffing, and password reuse, has exposed significant weaknesses in password-only authentication systems, prompting the development of more advanced techniques such as multi-factor authentication and biometric authentication. [1,2] This study assesses password authentication techniques for enhancing software security through a comparative analysis of password-based authentication, multi-factor authentication, and biometric authentication. A qualitative comparative research design was adopted using secondary data obtained from textbooks, peer-reviewed journal articles, conference proceedings, and international standards, with techniques evaluated on security effectiveness, usability, implementation cost, reliability, and resistance to cyberattacks. [3,5] Findings indicate that although password-based authentication remains essential because of its widespread acceptance, combining it with additional authentication factors significantly improves software security and reduces the risk of unauthorized access. The study concludes that organizations should adopt multi-layered authentication mechanisms to achieve stronger protection for modern software applications.

DeepSecAuth: Multi-Factor Authentication Framework Using Behavioral And Facial Biometrics

Authors: Mrs. Aruna C, S. Sumalatha

Abstract: In this day and age, cyberattacks have become increasingly sophisticated such that traditional authentication mechanisms like passwords can no longer secure sensitive information. In this paper, DeepSecAuth is proposed, an advanced authentication scheme that integrates keystroke dynamics behavioral biometrics with deepFace-based face recognition technology with the help of a honeypot deception technique. DeepSecAuth uses the one-class support vector machine algorithm to learn the unique typing pattern of the user at the time of registration, achieving 92% verification accuracy while the DeepFace achieves 95% recognition accuracy as the second level of authentication. The experimental results show that the integrated system has achieved 97% overall authentication accuracy with FRR of 1.2% and FAR of 1.6%, which is far better than the unimodal authentication schemes. The honeypot technique ensures that the credentials theft and spoofing attack are mitigated.

DOI: http://doi.org/10.5281/zenodo.21675751

Zero-Trust Architecture Framework For Advanced Threat Detection And Mitigation In Enterprise Networks

Authors: S.Janani, S.Vedavalli

Abstract: Perimeter-based security paradigms have proved insufficient in defending enterprises against sophisticated cyberattacks like Advanced Persistent Threats (APT), insider threat misuse, and supply chain attacks. Zero Trust Architecture (ZTA) offers a reliable solution to such cyberattacks, relying on continuous verification of users, devices, and network activity by following the "Never trust, always verify" paradigm. This paper introduces an elaborate ZTA framework that integrates SIEM, SOAR, and UEBA systems to detect and mitigate advanced threats. The framework utilizes machine learning techniques such as Isolation Forest for detecting behavioral anomalies and dynamic risk scoring. Multi-layered enforcement is implemented at three different levels of Identity, Device, and Network enforcement. Experiments conducted prove the effectiveness of this framework with Mean Time to Respond (MTTR) less than 10 seconds for critical threat cases and an accuracy of more than 95% for detecting anomalies in behavior patterns.

DOI: http://doi.org/10.5281/zenodo.21675846

Cloud Computing Adoption And Its Impact On Organizational Performance And Scalability

Authors: Nerella Paul Ranjit Kumar, Venkata Phani Rajesh Neelamraju

Abstract: The phenomenon of cloud computing has turned out to be revolutionary for the functioning of contemporary businesses due to numerous advantages associated with it. In the present research paper, the impact of cloud computing implementation on organizational performance is analyzed with particular consideration being given to scalability as a mediator. With the help of the extensive literature review and the use of quantitative methods, the positive influence of cloud computing implementation on such organizational characteristics as efficiency, profitability, and strategy formation is proved. At the same time, it is indicated that there are considerable obstacles related to such aspects as privacy issues, difficulties in managing costs, and integrational problems. The innovative methodology presented in the research includes the introduction of the Cloud Resource Optimization Framework (CROF) based on Learning Automata scheduling. According to the empirical results, organizations using well-designed approaches to cloud computing experience 46.5% improvement in resource utilization and 41.9% increase in decision accuracy.

DOI: http://doi.org/10.5281/zenodo.21675938

AI-Driven Financial Fraud Detection Framework For Digital Payment Ecosystems

Authors: Dr. Gundupagi Manjunath, Dr. N Chandan Prashad

Abstract: The increasing digitization of financial services has led to the transformation of payment systems through improved convenience but also increased vulnerability to fraudulent attacks. Rule-based approaches to detecting such threats, limited by their static nature and preconceived threshold limits, prove inadequate to combat evolving threat models. In this paper, we propose a new approach to fraud detection through artificial intelligence that utilizes ensemble machine learning and deep neural networks to create a zero-trust architecture security mesh. Our system utilizes Random Forest and XGBoost classifiers along with LSTM sequence aware networks and adaptive learning capability for addressing concept drift. The experiments conducted on our algorithm based on the PaySim synthetic transaction data set demonstrate significantly better fraud detection performance, with an accuracy of 99.96%, precision of 91.84%, and recall of 89.12%. Our approach lowers false positive rates by 21% compared to traditional gradient boosting baselines while sustaining sub-120ms inference time.

DOI: http://doi.org/10.5281/zenodo.21676072

AI-Driven Intrusion Detection System For Smart Network Security

Authors: Guru Angel Daisy M

Abstract: Due to the growing number and level of sophistication of cyber attacks, there is an urgent need for intelligent, adaptive and automated systems for detection and reaction on the threats. In this paper we propose a new Artificial Intelligence based Intrusion Detection System (AI-IDS), which uses a hybrid architecture with integration of ensembles of machine learning and deep learning algorithms in order to increase network security. Our method uses Random Forest, Support Vector Machine and Convolutional Neural Network algorithms to analyze traffic data and detect the normal traffic and the U2R attack with frequency less than one percent with accuracy of 98% and 94.3% respectively. According to our experiments on the NSL-KDD dataset, the performance of our system is significantly higher with false positive rate equal to 0.001%, precision equal to 0.99 and real time detection latency of 0.2 milliseconds. Multi class classification in our system helps us to classify DoS, Probe, R2L and U2R attacks much better than signature-based solutions do.

DOI: http://doi.org/10.5281/zenodo.21676208

Recent Advances in Mechanical Engineering Design Optimization Using Artificial Intelligence

Authors: Mr. Rahul Khobragade, Mr. Rahul Ghotkar, Ms. Amisha Malviya

Abstract: Mechanical engineering design has evolved significantly over the past decade due to the rapid development of Artificial Intelligence (AI) technologies. Traditional engineering design methods primarily depend on analytical calculations, empirical knowledge, numerical simulations, and repeated trial-and-error processes, which often require considerable time, computational effort, and engineering expertise. As engineering systems become increasingly complex and performance requirements continue to rise, conventional optimization techniques face limitations in handling multi-objective design problems, nonlinear constraints, and large-scale design spaces. Artificial Intelligence has emerged as a transformative technology that enables intelligent, adaptive, and data-driven design optimization capable of improving product performance, reducing development time, minimizing manufacturing costs, and enhancing sustainability. Consequently, AI-based optimization has become an essential component of modern mechanical engineering design. Recent advances in Artificial Intelligence have introduced numerous intelligent algorithms capable of solving complex engineering optimization problems more efficiently than traditional approaches. Technologies such as Machine Learning (ML), Deep Learning (DL), Artificial Neural Networks (ANN), Genetic Algorithms (GA), Particle Swarm Optimization (PSO), Reinforcement Learning (RL), Fuzzy Logic, Evolutionary Algorithms (EA), and Digital Twin technology are increasingly integrated with Computer-Aided Design (CAD), Computer-Aided Engineering (CAE), and Finite Element Analysis (FEA) to create intelligent engineering design environments. These technologies enable engineers to explore a large number of design alternatives automatically while simultaneously satisfying performance, reliability, cost, manufacturability, and environmental constraints. Artificial Intelligence significantly enhances the engineering design process by enabling intelligent prediction, optimization, automation, and decision support. Machine Learning models learn from historical engineering data to predict product behaviour, structural performance, material properties, manufacturing outcomes, and system reliability. Deep Learning techniques process complex multidimensional engineering datasets, allowing rapid evaluation of design alternatives without repeatedly performing computationally expensive simulations. Artificial Neural Networks provide accurate approximations of engineering responses, reducing computational time during optimization while maintaining high prediction accuracy.

DOI: https://doi.org/10.5281/zenodo.21700267

Efficient Deployment of Transformer-Based Large Language Models on Edge Computing Devices: Strategies, Challenges, Optimization Techniques, and Future Research Directions

Authors: Research Scholar Mr.K.Raju, Assistant Professor Dr.Rahul Kumar Budania

Abstract: A fast developing field of artificial intelligence research involves the nexus of edge computing, Transformer architecture, and Large Language Models (LLMs). Because of their high processing and memory needs, LLMs cannot be directly implemented on low-resource edge devices, despite their remarkable capacity to understand and produce natural language. The implementation of Transformer-based LLMs on edge computing platforms is examined in this study, with an emphasis on methods such as system-level approaches, inference optimization, model compression, and architectural improvements. The main issues with edge devices' constrained power, memory, and CPU capabilities are also covered. This work also looks at new compact LLMs and optimization techniques that increase deployment efficiency, lower computational costs, and enhance data security and privacy. A list of upcoming research challenges to enable effective, scalable, and intelligent edge-based AI applications concludes the survey.

DOI: https://doi.org/10.5281/zenodo.21701699

Deploying Large Language Models on Edge Computing Devices: Architectures, Optimization Techniques, Challenges, and Future Directions

Authors: Research Scholar Mr.K.Raju, Assistant Professor Dr.Rahul Kumar Budania

Abstract: By providing sophisticated natural language interpretation, reasoning, and decision-making capabilities, large language models (LLMs) have transformed artificial intelligence (AI). However, many real-time and mission-critical applications cannot profit from the standard cloud-based deployment of LLMs due to concerns such high communication latency, excessive bandwidth utilization, privacy issues, and reliance on constant network connectivity. By moving AI processors closer to data sources and enabling intelligent processing directly on edge devices while lowering reaction times, boosting system autonomy, and improving privacy, integrating LLMs with Edge Intelligence (EI) gets around these restrictions. By looking at the most recent architectural frameworks, deployment techniques, and learning paradigms created for resource-constrained edge contexts, this survey offers a thorough analysis of LLM-based Edge Intelligence. To facilitate the effective deployment of LLMs across different edge infrastructures, it covers cloud-edge, edge-only, cloud-edge-client, federated learning, peer-to-peer learning, and knowledge distillation techniques. Advanced optimization strategies like model compression, quantization, pruning, computation offloading, effective memory management, edge caching, and lightweight Small Language Models (SLMs), which dramatically lower computational overhead while preserving high inference performance on resource-constrained devices, are also covered in the survey. The report highlights numerous real-world applications of LLM-powered Edge Intelligence, including software engineering, robotics, intelligent transportation systems, intelligent healthcare, autonomous driving, the Industrial Internet of Things (IIoT), upcoming 6G communication networks, and smart cities. The survey highlights the significance of creating transparent, moral, and reliable AI systems that adhere to new moral and legal standards. Scalable LLM deployment, effective resource use, energy-conscious computing, intelligent model adaptation, distributed learning, secure edge collaboration, and smooth interaction with future 6G and beyond communication networks are just a few of the subjects covered in the survey's conclusion. It also describes future research directions and highlights existing research shortages. Overall, because it offers a comprehensive analysis of the architectures, optimization techniques, applications, security issues, and potential future advancements of LLM-based Edge Intelligence, the paper is a useful tool for researchers, practitioners, and developers. This enables the creation of effective, safe, scalable, and reliable edge computing systems driven by AI.

DOI: https://doi.org/10.5281/zenodo.21701918

Artificial Intelligence-Driven Pharmacovigilance 5.0: A Systematic Review and Future Framework for Predictive Drug Safety Surveillance Using Machine Learning and Digital Health Technologies

Authors: Asila Sallauddin Shaikh

Abstract: Pharmacovigilance is a critical component of healthcare systems responsible for ensuring medicine safety through detection, assessment, understanding, and prevention of adverse drug reactions. Conventional approaches based on spontaneous reporting and manual evaluation face challenges including under-reporting, delayed signal detection, fragmented healthcare information, and limited predictive capability. This review explores the transformative role of Artificial Intelligence (AI) in Pharmacovigilance 5.0 by integrating machine learning, deep learning, natural language processing, predictive analytics, and real-world evidence. AI-driven systems have demonstrated potential in adverse drug reaction detection, safety signal identification, risk prediction, and regulatory decision support. The manuscript discusses emerging opportunities, technological frameworks, regulatory challenges, and future perspectives for establishing intelligent, predictive, and patient-centric drug safety surveillance systems.

DOI:

Deep Learning-Based Intelligent Bank Cheque Verification Using Advanced Image Processing Techniques

Authors: Assistant Professor G. Sudheer kumar, K. Bhagyasree

Abstract: Bank cheque verification is a critical banking operation that requires high accuracy to minimize fraud and processing delays. This study introduces an intelligent cheque verification framework that combines image processing with deep learning techniques to automate the authentication process. The proposed approach extracts essential cheque details such as the IFSC code, cheque number, account number, handwritten amount, printed text, and account holder's signature from scanned cheque images. Image preprocessing techniques are applied to enhance image quality and isolate important regions before analysis. Optical Character Recognition (OCR) is employed to recognize printed characters, while a Convolutional Neural Network (CNN) is utilized for identifying handwritten numerical values. Signature verification is performed using Scale-Invariant Feature Transform (SIFT) for feature extraction and a Support Vector Machine (SVM) for classification. The integrated framework improves verification reliability, accelerates cheque processing, and reduces manual intervention. Experimental evaluation on a standard cheque image dataset demonstrates that the proposed system achieves high recognition accuracy for text, handwritten digits, and signatures, making it a practical solution for secure and efficient banking applications.

DOI: https://doi.org/10.5281/zenodo.21739340

An Explainable Hybrid Deep Learning Framework for Automated Knee Osteoporosis Diagnosis Using CNNs and Fine-Tuned Vision-Language Models

Authors: Jonnalagadda Somaiah, Professor Dr. Koppula Chinabusi

Abstract: Knee osteoporosis is a progressive bone disorder that increases fracture risk and significantly reduces mobility, making early diagnosis essential for effective treatment. Conventional deep learning approaches based solely on Convolutional Neural Networks (CNNs) often struggle to capture complex clinical patterns and contextual information present in medical data. This study proposes an explainable hybrid deep learning framework that integrates CNN-based feature extraction with fine-tuned Vision-Language Models (VLMs) to improve the automated classification of knee osteoporosis from X-ray images. Multiple CNN architectures, including ResNet50, DenseNet121, MobileNetV2, InceptionV3, Xception, and NASNetMobile, are evaluated and compared with a multimodal transformer-based model combining Vision Transformer (ViT) and BERT. The proposed framework incorporates transfer learning, multimodal feature fusion, hyperparameter optimization, and Grad-CAM-based visual interpretation to enhance diagnostic accuracy and model transparency. Experimental evaluation demonstrates that the fine-tuned multimodal model achieves superior classification performance with an accuracy of 93%, outperforming conventional CNN models while providing clinically interpretable predictions. The proposed framework offers a reliable, scalable, and explainable computer-aided diagnosis system that can support healthcare professionals in the early detection and severity assessment of knee osteoporosis.

DOI: https://doi.org/10.5281/zenodo.21739491

Machine Learning based Blackhole Attacker Detection and Prevention Approach in VANET

Authors: Kashifa Khan, Associate Professor Devdas Saraswat

Abstract: Vehicular Ad hoc Network (VANETs) face numerous issues due to the free movement of vehicles within the network. These problems include high bandwidth consumption, as vehicles often wait for responses and properly received traffic information from one another. The innovative RSML system enhances communication speed but requires additional security features to identify and shield the VANET from blackhole attacks. RSML incorporates machine learning techniques within roadside units (RSUs) to safeguard the transportation network. In the context of routing, blackhole attackers seek to disrupt normal routing functions. These attackers consistently generate high sequence numbers to participate in the routing process. The RSML method utilizes machine learning to effectively detect and mitigate black hole threats in vehicular ad hoc networks. By employing advanced algorithms, the RSML method can adapt to changing attack patterns, ensuring strong protection for vehicle communications. Establishing a route in vehicular communication is challenging because a malicious vehicle can exploit this process for its advantage. The RSML approach maintains the record of attacker malicious functioning and match the malicious functions. Therefore, it is essential to select a routing technique that guarantees secure communication. The performance of RSML is compared with existing balckhole attack, T-AODV and ANN-AODV. The performance of novel RSML scheme is better.

DOI: https://doi.org/10.5281/zenodo.21768288

A Novel Flooding based Heuristic Security Approach for Vampire Attack in FANET

Authors: Chhatrapani Gautam, Associate Professor Dr. Ankur Pandey

Abstract: FANET Flying ad hoc networks are a promising technology and have enormous potential to be working in critical situations in farming, land investigation, habitat monitoring, smart homes, and many more scenarios. One of the major challenge in FANET facing today is security. In this paper, we propose a flooding identification-based heuristic scheme against vampire attackers in FANET. These types of attacks flood the network with unnecessary packets, consuming bandwidth and affecting data delivery due to high energy consumption. The main aim is to visualize the effect of a vampire attack on the network and identify the UAVs that are affecting its performance. The heuristic security scheme checks the profile of each UAV in the network, and if an attacker is identified as one of the UAV or UAVs that flooding unnecessary packets and the security scheme has blocked that attacker's malicious functioning. The functioning of heuristic based flooding scheme has blocked the performance of the attacker. The attacker aim is only to disturb the normal communication among the UAVs in FANET. The performance of a network is measured on the basis of performance metrics such as routing load, throughput, and end to end delay. The simulation results indicate that the performance is the same for both normal routing and the heuristic scheme, demonstrating that the security scheme is effective and shows better performance. This effectiveness highlights the robustness of the security scheme for vampire attackers in maintaining network integrity, even in the face of potential threats.

DOI: https://doi.org/10.5281/zenodo.21768451

Influence of Social Media on Eating Habits: A Comparative Study between Boys and Girls (12–18 Years)

Authors: P. Jahnavi, Dr. P. Ashlesha

Abstract: Background: The widespread use of social media has transformed the way adolescents access information related to food, nutrition, and body image. Continuous exposure to food-related content, advertisements, and peer interactions on digital platforms may influence dietary behaviours and lifestyle choices during adolescence. Objectives: To evaluate the influence of social media on eating habits among adolescents aged 12–18 years and to compare its effects on food preferences, body image perception, and dietary behaviours between boys and girls. Methods: A comparative cross-sectional study was conducted among 100 adolescents (50 boys, 50 girls) aged 12–18 years using a structured questionnaire. Information regarding social media usage, food preferences, eating behaviours, body image perception, dieting practices, and peer influence was collected and analyzed. Associations between gender and study variables were assessed using the Chi-square test in Microsoft Excel, with statistical significance considered at p<0.05. Results: Social media demonstrated a considerable influence on the eating behaviours of adolescents. Boys were more likely to report positive responses to healthy eating-related content, whereas girls showed greater susceptibility to body image and dieting messages shared through social media. Most gender-based differences were not statistically significant. A significant association was observed only for perception of healthy versus unhealthy eating habits (p=0.026). No significant associations were identified for dieting behaviour (p=0.537), food preferences (p=0.343), peer influence (p=0.368), or exposure to social media food-related content (p=0.951). Conclusions: Social media is an important determinant of adolescents' dietary behaviours irrespective of gender. Incorporating nutrition education and media literacy into adolescent health programmes may enable young people to critically evaluate online information, make healthier food choices, and adopt sustainable dietary practices.

DOI: https://doi.org/10.5281/zenodo.21768947

Mathematics & Artificial Intelligence for Industry from Small Business to Large Enterprises

Authors: Dr Pawan Kumar, Anil Kumar, Dr Ram Bhushan, Dr Rajendra Kumar Tripathi, Ajit Shukla, Mrs.Madhu Sonkar, Dr Shivesh Mani Tripathi

Abstract: The convergence of mathematics and artificial intelligence is fundamentally transforming how industries operate; however, the pace and nature of this transformation vary considerably across organizations of different sizes. This article explores how mathematical optimization techniques, machine learning algorithms, and AI systems can generate competitive advantages, while critically examining the asymmetric barriers hindering their adoption in organizations. We propose a conceptual framework demonstrating that while large enterprises possess the resources to develop customized AI solutions, small and medium-sized enterprises (SMEs) are increasingly leveraging available alternatives, such as AI service subscriptions, conversational interfaces that mask mathematical complexities, and open-source optimization engines. Based on contemporary research and emerging industry practices, the analysis underscores that mathematical accuracy embedded in AI infrastructure remains essential for reliable industrial deployment. The study concludes that increased accessibility to mathematical AI resources, coupled with system designs that clearly distinguish between cognitive reasoning and numerical processing, represents a viable path toward widespread industrial modernization.

DOI: https://doi.org/10.5281/zenodo.21785297

Design, Modeling and Performance Evaluation of a Dual-Axis Solar Tracking Photovoltaic System Using MATLAB/Simulink

Authors: Premdas T P, Dr. Rajeesh C R

Abstract: Stationary photovoltaic (PV) installations suffer significant energy losses because the panel surface is rarely perpendicular to the incoming solar beam, a phenomenon governed by cosine losses that intensify during morning and evening hours. This paper presents the design, mathematical modeling and simulation-based performance evaluation of a dual-axis sun tracking PV system developed in MATLAB/Simulink. An astronomical, algorithm-based control strategy computes the reference azimuth and elevation angles from geographic latitude, day-of-year and local solar time, while an error-driven, dead-band controller actuates the tracking motors to keep the incidence angle close to zero throughout the day. The single-diode equivalent-circuit model, together with temperature-corrected electrical parameters, is used to translate the tracked irradiance into instantaneous power. Simulation results for a representative 250 W module show that the dual-axis tracker sustains a broad, flat power plateau compared with the narrow bell-shaped curve of a fixed-tilt panel, yielding an energy gain of approximately 30-40%. Field-comparable data from an identical pair of 1.3 kWp systems corroborate the simulation, showing a 36.47% annual energy improvement with the simulated and measured values differing by less than 5%. An economic assessment indicates that, despite a higher upfront cost, the additional energy revenue offsets the investment within a comparable payback period to fixed systems, while a carbon-mitigation analysis confirms measurable emission-reduction benefits. The study concludes that dual-axis tracking is most advantageous for land-constrained, high-yield and research-oriented deployments, whereas fixed systems remain preferable for small residential installations.

DOI: https://doi.org/10.5281/zenodo.21785774

A Fine-Tuned Vision Transformer with Grad-CAM Explainability for Real-Time 36-Class Waste Classification and REST API Deployment

Authors: Gurdarshan Singh, Shivam, Yash Singh

Abstract: Sorting waste by hand is slow, error-prone, and still one of the biggest bottlenecks in recycling streams. This paper presents WasteVision AI, an end-to-end system for fine-grained waste classification built around a Vision Transformer (ViT-Base/16), fine-tuned on the 36-class TriCascade WasteImage dataset (35,264 images) through a two-stage transfer-learning schedule. The backbone is first warmed up as a whole with a frozen classifier head under a class-weighted, label-smoothed cross-entropy loss, after which the last four transformer blocks are unfrozen and fine-tuning continues. On a held-out test set of 3,527 images, the resulting model reaches 95.66% top-1 accuracy, 98.27% top-3 accuracy, and a macro-averaged F1-score of 0.93. To explain its predictions, we adapt Grad-CAM to work on the transformer’s patch-token hidden states, producing pixel-wise heatmaps that localise the discriminative regions of each waste item; predictions below 50% confidence are flagged “Unknown Waste” rather than forced into a class. The whole pipeline is exposed through a lightweight Flask REST API that returns the predicted class, a calibrated confidence score, and a base64-encoded Grad-CAM overlay in a single call. Against a recent three-stage DP-CNN/Ensemble-ELM benchmark evaluated on the same dataset, our single-stage ViT pipeline gains over 10 percentage points of fine-grained (36-class) accuracy, though at a higher parameter cost — a trade-off worth weighing when choosing between transformer- and CNN-based deployments for waste sorting.

Improvememnt of Post-Fire Mechanical Properties of Concrete Using Calcinated Kaolin (Metakaolin)

Authors: Engr. Ezea Boniface, Olorunshola Banjo, Fadiran David, Engr. Dr. Ugwuoke Malachy Okonkwo, Dr. J O Labiran

Abstract: Fire exposure significantly alters the mechanical and durability properties of concrete, often leading to substantial reductions in structural capacity. This study investigates the improvement of residual mechanical properties of concrete exposed to elevated temperatures through partial replacement of ordinary Portland cement with calcined kaolin (metakaolin, MK). Concrete specimens were produced with 0%, 2.5%, 5%, 7.5%, and 10% metakaolin replacement by weight of cement. The water-cement ratio was varied appropriately to maintain workability as MK content increased. Specimens were cured for 28 and 56 days and subsequently exposed to heat loads of 200°C, 400°C, 600°C, and 800°C. Residual compressive strength, mass loss, and water absorption were determined to evaluate post-fire performance. Results indicate that compressive strength improved at ambient conditions with increasing MK content, with optimum performance observed between 7.5% and 10% replacement. At elevated temperatures, MK-modified concrete demonstrated superior residual strength retention compared to control specimens, particularly at 600°C and 800°C. Mass loss and water absorption increased with temperature across all mixes; however, mixes containing higher MK percentages exhibited comparatively reduced deterioration, attributed to pore refinement and reduced calcium hydroxide content. Specimens cured for 56 days consistently showed better residual performance than 28-day specimens, highlighting the importance of hydration maturity in thermal resistance. Correlation analysis revealed an inverse relationship between water absorption and residual compressive strength, confirming that microstructural densification significantly enhances fire resistance. The findings support the use of metakaolin as an effective supplementary cementitious material for improving the thermal stability and post-fire structural integrity of concrete. The study contributes to performance-based fire design by emphasizing material performance enhancement rather than reliance solely on prescriptive measures such as concrete cover thickness.

DOI: https://doi.org/10.5281/zenodo.21837294

Predictive AI-Based Battery cooling and Thermal control system for EV Vehicles: A Literature Review

Authors: Associate Professor H R Patil, Associate Professor M M Ganganallimath, Manu G. Hunagund, Naveen Itagi, Prajwal P. Kajagar, Prashant Y. Ghorpade

Abstract: The necessity for effective Battery Thermal Management Systems (BTMS) to keep lithium-ion batteries within their ideal temperature range has grown due to the explosive growth of electric vehicles (EVs). Battery longevity, safety, charging efficiency, and overall vehicle performance are all strongly impacted by battery temperature. Although they offer efficient thermal management, conventional cooling techniques including air cooling, liquid cooling, phase change materials, and heat pipes are constrained in dynamic driving and fast-charging scenarios. Predictive battery temperature management employing machine learning, deep learning, reinforcement learning, and Model Predictive Control (MPC) has been made possible by recent developments in artificial intelligence (AI). These methods increase battery safety, minimize energy usage, optimize cooling procedures, and precisely anticipate battery temperature. Conventional cooling technologies, AI-based thermal prediction techniques, predictive control tactics, and current advancements in intelligent BTMS are all summarized in this paper. Additionally, it draws attention to present issues and potential paths, such as Edge AI, Physics-Informed AI, and Digital Twins. All things considered, predictive AI-based thermal management presents a viable way to improve battery safety, energy economy, and next-generation electric vehicle performance.

DOI: https://doi.org/10.5281/zenodo.21847617

A Systematic Analysis of the Surge in Petrol and Diesel Prices in India Since 2013 An Using by Robust Fuzzy Regression Estimator

Authors: Assistant Professor P. Anandhi, Assistant Professor S.Gunasekaran, Assistant Professor S.Kavithanjali, Assistant Professor M. Srividya

Abstract: India has emerged as the third base-prime trader of crude oil globally, superior previous benchmarks in 2020 and fulfilling completed 84% of its oil needs done imports. This investigates the reasons behind the sharp increases in petrol and diesel prices and their subsequent impact on India's economy. The rapid escalation in crude oil prices exerts significant pressure on the economy, driven by factors such as fluctuating supply and demand, OPEC policies, high government taxes on petroleum products, and the rupee-dollar exchange rate. To address these issues, this paper employs the Robust Fuzzy Regression Estimator (RFRE) to identify and mitigate irregular data points, thereby refining the analysis of price trends. The study finds that rising energy prices contribute to increased manufacturing costs, lower GDP growth, a widening current account deficit (CAD), tax variations, and heightened costs for transportation and essential goods. Contrary to the belief that global crude oil prices are the primary driver of domestic fuel prices, the analysis concludes that the economic policies of both central and state governments play a more critical role. This research provides valuable insights into the complexities of fuel pricing and offers a clearer understanding of the statistical economic implications of energy price volatility.

Snacking Pattern Among Teenagers in Hyderabad

Authors: Sameera Begum, P Ashelsha

Abstract: With rapid urbanization, the traditional dietary practices of Indian adolescents are shifting toward convenience-driven, processed, and energy-dense snacks. The present study investigates the snacking patterns, lifestyle behaviours, and dietary habits of teenagers aged 16–19 years in Hyderabad. These changes are influenced by various factors including social media, digital food platforms, peer pressure, and academic schedules. The research highlights a growing dependence on snacks such as chips, chocolates, fast food, and sugary beverages, which often replace nutritious meals. Data was collected using structured questionnaires from 100 students selected via random sampling from two institutions. The study assessed socio-demographic profiles, anthropometric data, food frequency, clinical history, and health-related outcomes. The findings revealed that while 57% of participants had normal BMI, 32% were underweight, and 11% fell into overweight or obese categories. Many teenagers reported skipping meals and lacking proper snack timing, with only 39% following structured snacking habits. A majority of students (64%) used online food delivery apps, showing a preference for processed and fast foods. Health issues like poor digestion, menstrual irregularities, and weight gain were frequently linked to unhealthy snacking. Despite some awareness, many participants were unclear about the nutritional impact of their choices. The study also explored the influence of family habits, physical activity, and water intake on dietary behaviour.

DOI: http://doi.org/10.5281/zenodo.21850781

Digital Transformation in Manufacturing: A Framework for Oracle Fusion Cloud Financials Adoption Using Modern Integration Platforms

Authors: Utkarsh Mehta

Abstract: Digital transformation in manufacturing organizations require a unified financial, supply chain, and operational backbone. Oracle Fusion Cloud Financials has emerged as a leading ERP platform for organizations seeking modernization, resilience, and real time decision making. This article presents a structured framework for adopting Fusion Financials using cloud native integration platforms such as Oracle Integration Cloud (OIC), Google Cloud Platform (GCP), and BigQuery (GBQ) based on real-world scenario. During Oracle Fusion Cloud Financials implementation, the Oracle Fusion Financials Integration is considered to be a critical component in the whole enterprise architecture where financial data (meta data and transaction) must seamlessly flow between multiple systems such as Order Management System (OMS), CRM, payroll, banking platforms, legacy ERPs, and third-party applications. During implementation lifecycle, integrations around Financials modules like General Ledger, Receivables, Payables, Purchasing, Cash Management will ensure data consistency, compliance and operation efficiency through data pipeline automation [1]. Oracle Fusion Financials Integration refers to the process of connecting Oracle Financial modules with external or internal systems to exchange financial data automatically.

Role of Dinacharya, Ritucharya, and Sadvritta in Public Health Promotion: An Evidence-Based Review

Authors: Dr. Vikas Dubey

Abstract: Dinacharya (daily regimen), Ritucharya (seasonal regimen), and Sadvritta (code of ethical/behavioral conduct) constitute the three principal lifestyle-based pillars of Swasthavritta, the Ayurvedic discipline of preventive and social medicine. This review examines the convergence between these classical constructs and contemporary biomedical evidence in chronobiology, seasonal metabolic/immune physiology, and behavioral/positive psychology. Dinacharya's alignment with circadian biology is the best-substantiated of the three, supported by molecular chronobiology (201 7 Nobel Prize-recognized clock genes) and clinical chronodisruption data. Ritucharya's seasonal-adaptation framework corresponds to documented seasonal variation in gene expression, gut microbiome composition, and metabolic markers, though direct outcome trials specific to Ritucharya remain limited. Sadvritta's ethical-conduct framework parallels established constructs in positive psychology (prosocial behavior, self-regulation, gratitude, compassion) with plausible but under-quantified mental-health impact. Collectively, these three pillars represent a coherent, low-cost, non-pharmacological framework for public health promotion and NCD prevention, though the evidence base is dominated by narrative/conceptual reviews rather than controlled outcome trials — a gap this paper identifies as the priority for future research.

DOI: https://doi.org/10.5281/zenodo.21868975

Sustainable and Smart: The Role of AI and Automation in Greening Aviation Logistics

Authors: Ms.Tincy Susan Baby

Abstract: Background- The aviation industry facilitates global mobility and trade and at the same time, it only accounts for 2-3 percent of the world’s carbon emissions. With the growing industry demand, aviation companies find themselves in a tight spot—striking a balance between operational efficiency and true sustainability. Objective- This review aims to address emissions and operational performance enhancement in aviation logistics with the help of Artificial Intelligence (AI) and automation technologies. Methods-This paper integrates the application of machine learning, computer vision, natural language processing, robotics, and industrial process automation in aviation industry with the help of case studies from India and the rest of the world alongside industry white papers and academic literature. It takes a systematic literature review of academic studies, white papers from the industry and documented implementations to assess the emerging role and impact of AI-driven technologies in aviation. Result- The findings indicate that AI applications can leverage predictive analytics, optimize resource utilization, enable autonomous operations, and usher in a new passenger experience. Qualitative evidence suggests significant improvements in operational efficiency, system resilience, and service consistency on the part of leading industry players. Conclusion- Collectively, these implementations help the aviation industry carbon reduction targets, including the IATA net zero 2050 emission target. Nonetheless, the more flexible AI systems face integration difficulty, high capital investment, and employee preparedness. The review determined that AI and automation could expedite the shift towards climate-friendly, intelligent aviation logistics, provided strong policies, quality data, and collaboration among all stakeholders is employed.

DOI: https://doi.org/10.5281/zenodo.21869080

Comparative Study of Atmospheric Corrosion of Carbon Steel in Gas-Flaring and Non-Gas Flaring Environments of the Niger Delta

Authors: Titus Ndiele Amadi

Abstract: This study presents a comparative assessment of atmospheric corrosion of carbon steel exposed in gas-flaring and non-gas flaring environments of the Niger Delta. The investigation examines corrosion mechanisms, and corrosion rates, and its implications for industrial infrastructure. Results indicate that carbon steel exposed in gas-flaring areas experiences substantially higher corrosion rates than steel exposed in non-gas flaring areas due to increased pollutant concentrations, acidic precipitation, elevated temperatures, and higher atmospheric conductivity. The study highlights the need for improved corrosion management strategies and stricter environmental controls in oil-producing regions.

DOI: https://doi.org/10.5281/zenodo.21872007

Performance Evaluation and Application of Discrete Time Queueing Models for Computer and Communication Networks

Authors: Pushpandra Kumar

Abstract: In this article, the study of discrete time queueing models for computer communication networks and their applications to queueing models. It is expected that a discrete time model is "slotted". During a slot, a discrete time queue may accept and service a maximum of one packet each. The major reason of discrete time queueing systems have attracted interest of potential applications that take use of slotted time. Comparing how continuous time models are handled, we take into consideration several conclusions inside a single discrete-time framework. This method has two benefits: it parallels the treatment of continuous time models as much as feasible, and It streamlines the handling of several discrete-time models under a single structure. This enables us to compare and contrast the continuous and discrete time queues. Specifically, Our applications related to birth and death center on the discrete-time single server model with bernoulli arrivals and geometric service times. We give the reader a straightforward, meticulous analysis covering every one of the five scheduling guidelines examined inside the written works, focusing especially on the stationary distributions at the pre arrival, slot centers, and slot margins.

DOI: https://doi.org/10.5281/zenodo.21900298

Comprehensive Review On Fuel Cells and Renewable Hydrogen

Authors: Adeleye, S. A., Olurunshola. O. Z., Ipindola O., Oriowojide R. P.

Abstract: Fuel cells and renewable hydrogen technologies are key enablers for the global transition to carbon neutrality and the United Nations Sustainable Development Goal 7 (SDG 7) on affordable, reliable, sustainable and clean energy for all. The status, engineering challenges, and future prospects of these technologies are reviewed herein. This work is motivated by the imperative to reduce our reliance on fossil fuels, which contribute about 73% of global greenhouse gas emissions and are inconsistent with international climate goals. There has been a lot of investment in solar and wind energy, but sectors that are hard to decarbonize, such as heavy industry, long-haul transportation, maritime shipping and seasonal grid storage, continue to pose decarbonization challenges. Fuel cells and renewable hydrogen are seen as promising solutions for these challenges. The review is a holistic narrative synthesis of peer-reviewed studies, institutional roadmaps, and techno-economic analyses published between 2000 and 2025. The article discusses the basic electrochemical principles of fuel cells and water electrolysis. The article discusses five main types of fuel cells: PEMFC, AFC, PAFC, MCFC, and SOFC. The article evaluates the efficiency, cost, durability, and applicability of fuel cells. We also discuss renewable hydrogen production pathways (electrolysis, thermochemical conversion, and biological methods), hydrogen storage and distribution technologies, and key material and engineering bottlenecks, against the backdrop of global energy transition policies and Nigeria’s energy access challenges under SDG 7. PEMFCs are preferred for transportation due to high power density and fast start-up while SOFCs are preferred for stationary combined heat and power applications, the review finds. Nevertheless, commercialization is still hampered by the high cost of green hydrogen production, low round-trip energy efficiency, membrane degradation, instability of platinum-group metal catalysts, iridium scarcity, and lack of refueling infrastructure. We highlight a key gap in the literature, where existing work is still fragmented and technology specific, and lacks a holistic framework for the comparative integration of hydrogen production, storage, distribution and end-use systems – especially in a sub-Saharan African context. The review is also limited by the absence of detailed geopolitical analysis, coverage of nuclear hydrogen production, safety regulatory assessment and full lifecycle evaluations of commercial systems. The paper concludes that renewable hydrogen and fuel cells are complementary and essential technologies for SDG 7 and Net Zero 2050 and highlights the need for advances in materials science, infrastructure building and cost reduction to enable sustainable large-scale deployment.

DOI: http://doi.org/10.5281/zenodo.21903201

Research on Propeller Application Configurations on Aircraft

Authors: BSc Dinh Van Quyen, MSc Trong Thuong Tran

Abstract: This paper studies the use of propellers on aircraft. For single-engine aircraft, the propeller is commonly installed at the nose, on the fuselage, or at the tail, with the nose-mounted configuration being the most optimal. Multi-engine aircraft generally use symmetrically arranged propellers on both sides of the wings to increase power and safety. Propeller aircraft offer high efficiency at low speeds and low fuel consumption, making them suitable for transport, training, sport aviation, UAVs, and short-range flight. Their main limitation is the decrease in efficiency at high speeds, so they are less suitable for high-speed missions.

DOI: https://doi.org/10.5281/zenodo.21914178

A Dual-Key Based RSA Cryptosystem

Authors: Vasudeo Patil, Arpit Vijay Raikwar

Abstract: Rivest–Shamir–Adleman (RSA) is a widely used public-key cryptosystem based on a single public-private key pair. This paper proposes a dual-key RSA cryptosystem that employs two public keys with their corresponding private keys, introducing additional key dependency into the encryption and decryption processes. The correctness of the proposed scheme is established through a formal mathematical proof and illustrated with a numerical example. The proposed scheme and conventional RSA are implemented in Java using Apache NetBeans IDE 25, and their computational performance is evaluated under identical parameters. The experimental results indicate that the proposed scheme achieves performance comparable to conventional RSA. The security analysis using ProVerif verifies message and private-key secrecy under a symbolic adversarial model. The proposed dual-key RSA cryptosystem represents a structured and efficient extension of the classical RSA scheme suitable for academic and experimental cryptographic studies.

DOI: http://doi.org/10.5281/zenodo.21914877

Quantum Leaps Through The Mathematical Language Of Next Generation Computing

Authors: Dr. C. Radhiya Devi, Dr. P. Umadevi, Dr. C. Mohana Priya

Abstract: Quantum computing represents one of the most significant paradigms shifts in the history of computation, promising to solve classes of problems that remain intractable for classical machines. Unlike classical computing, which is built on Boolean algebra and binary logic, quantum computing is fundamentally a mathematical discipline rooted in linear algebra, complex vector spaces, and probability theory. This article examines the mathematical structures that underpin quantum computation, including qubits, superposition, entanglement, unitary transformations, and quantum gates, and explores how these structures enable algorithms such as Shor's and Grover's to outperform their classical counterparts. The discussion further extends to quantum error correction, the challenges of decoherence, and the interdisciplinary convergence of mathematics and computer science that is shaping next-generation computing architectures. The article concludes with a reflection on open mathematical problems and the future trajectory of quantum information science.

DOI: http://doi.org/10.5281/zenodo.21915898

A Study On The Prevalence And Awareness Of Prediabetes Among Adult Women Aged 20-40 Years

Authors: A Shireesha, P Ashlesha

Abstract: Prediabetes is a metabolic condition characterized by elevated blood glucose levels that are not yet high enough to be diagnosed as diabetes but carry an increased risk of progression to type 2 diabetes and associated complications. This study aimed to assess the prevalence and awareness of prediabetes among adult women aged 20–40 years in the Koti region. A survey was conducted with a sample size of 100 participants selected through simple random sampling. Information was gathered through a structured questionnaire that encompassed socio-demographic characteristics, anthropometric data, lifestyle behaviors, family and clinical background, eating habits, and food consumption frequency. Findings revealed that 55% of the respondents had blood glucose levels above normal, indicating a high pre-diabetic prevalence. 43% failed to meet the recommended 150 minutes per week. Obesity was present in 23% of participants, and 42% were overweight, highlighting the role of body mass in disease risk. A total of 47% of the individuals who took part in the study reported having a family history of diabetes. Awareness was low, with only 30% acknowledging the importance of a reported balanced diet, and 43% eating after 8 PM daily. Regular intake of sugary drinks along with a limited consumption of whole grains was also noted. The study concludes that despite a relatively educated population, awareness and preventive practices related to prediabetes remain inadequate. These findings emphasize the urgent need for health education and interventions promoting lifestyle modifications to reduce the risk of prediabetes.

Mastering Interpersonal Communication: The Human Element in Collaboration

Authors: Preksha Jain

Abstract: Interpersonal communication is a fundamental component of effective collaboration in modern professional environments. While communication primarily involves the transmission of information, interpersonal communication extends beyond information exchange to include empathy, emotional awareness, trust, active listening, and meaningful human connection. This distinction is particularly important in workplaces where diverse individuals must collaborate, resolve disagreements, exchange feedback, and achieve shared objectives. This paper examines the conceptual difference between communication and interpersonal communication and discusses five essential elements of the interpersonal communication process: sender, message, channel, receiver, and feedback. It further presents ten strategic principles for improving workplace interactions, including clarity, avoidance of unnecessary jargon, appreciation, positive body language, active listening, consistency, healthy boundaries, constructive feedback, appropriate channel selection, and the use of open-ended questions. Drawing upon Stephen R. Covey's principle of seeking first to understand before seeking to be understood, the paper emphasizes the importance of empathy and emotional intelligence in professional relationships. The discussion concludes that effective collaboration depends not only on the accurate transfer of information but also on the quality of human relationships through which that information is exchanged.

Human Psychology in the Era of Artificial Intelligence: Human–AI Interaction, Trust, Emotion, and Cognitive Behavior

Authors: Gowri Gireesh, Dr Abhilash S.Vasu

Abstract: Artificial intelligence (AI) has moved from task-oriented automation toward systems that communicate with, advise, and collaborate with humans. Consequently, psychological factors such as trust, emotion, cognitive bias, perceived intelligence, anthropomorphism, and decision confidence have become central to AI adoption. This paper presents a human-centered framework connecting psychological characteristics with AI system characteristics and interaction context. A focused literature synthesis is used to examine emotion-aware AI, human–AI trust, explainability, cognitive behavior, and the emerging influence of generative AI. The paper identifies research gaps in longitudinal assessment, culturally diverse datasets, multimodal psychological measurement, and calibration of appropriate—not excessive—trust. A proposed methodology combines questionnaire-based psychological measures, controlled human–AI experiments, behavioral indicators, and machine-learning analysis. The framework can support the design of trustworthy, transparent, adaptive, and psychologically compatible AI systems for education, healthcare, workplaces, and consumer applications.

DOI: https://doi.org/10.5281/zenodo.21944312

Journalists’ Opinion on Artificial Intelligence and its Implications for News Credibility: A Case Study of Zanis Lusaka

Authors: Kalani Muchima, Clinton Masumba, Christine Mubanga

Abstract: This qualitative study investigated journalists' opinions on artificial intelligence (AI) and its implications for news credibility at the Zambia News and Information Services (ZANIS) Lusaka office. The research was guided by three objectives: to identify the AI tools currently used by ZANIS journalists and their purposes; to examine journalists' perceptions of AI's implications for news credibility; and to explore the ethical issues associated with AI use. The study adopted an interpretivist research philosophy and employed a single-case study design. Data were collected through semi-structured interviews with six journalists purposively selected from different role categories (political reporter, health reporter, business reporter, senior editor, digital editor, and news photographer). Thematic analysis was conducted following Braun and Clarke's six-phase framework. The theoretical framework integrated the Technology Acceptance Model (TAM), Diffusion of Innovations Theory, Media Credibility Theory, and Professional Identity Theory. The findings revealed that AI adoption at ZANIS Lusaka is universal but entirely informal. All six participants reported using AI tools, primarily for transcription, drafting, and research. However, this adoption occurred without formal organisational policy, training, or guidance. Journalists learned about AI tools through self-directed experimentation, peer recommendations, and online tutorials. The most significant finding was the complete absence of any formal AI policy at ZANIS, creating risks of inconsistent practices, accountability gaps, and ethical blind spots. Regarding perceptions of news credibility, participants experienced a fundamental tension between AI's efficiency benefits and concerns about accuracy, authenticity, and audience trust. Older journalists expressed greater credibility concerns than their younger colleagues. Participants disagreed on whether and how AI use should be disclosed to audiences, reflecting the broader "transparency dilemma" in the literature. Ethical concerns identified included bias and cultural inappropriateness of AI tools developed outside Africa, accountability gaps for AI-generated errors, deskilling of fundamental journalistic skills, job displacement fears, and emerging concerns about privacy, data security, and source protection. The study concludes that while ZANIS journalists are enthusiastically adopting AI tools for their efficiency benefits, this adoption is occurring without the policy frameworks, training, and ethical guidance necessary to protect news credibility. The study recommends that ZANIS urgently develop a formal AI policy, establish mandatory training programmes, clarify disclosure standards and accountability mechanisms, and provide guidance on data privacy and source protection.

DOI: https://doi.org/10.5281/zenodo.21974295

Spore Morphology of Selected Pteridophytes from the Southern Western Ghats, India: A Taxonomic Perspective

Authors: Ramseena. A

Abstract: Pteridophytes constitute an important component of the flora of the Western Ghats, one of the world's biodiversity hotspots. Spore morphology has long been recognized as a valuable taxonomic character for species identification and phylogenetic interpretation. The present study investigated the spore morphology of selected pteridophytes collected from the Southern Western Ghats of Kerala, India. Twenty-one species representing 18 genera and 15 families were collected from diverse forest habitats. Spores were examined using light microscopy (LM) and scanning electron microscopy (SEM) following acetolysis preparation. Morphological characters including spore size, shape, aperture type, exospore thickness, and ornamentation were recorded. The study identified two major spore types, namely trilete and monolete spores, with considerable interspecific variation in ornamentation and dimensions. SEM analysis revealed diagnostic exospore patterns such as verrucate, reticulate, cristate, and granulose ornamentation that were not fully distinguishable under LM. The findings demonstrate that spore morphology provides reliable taxonomic evidence for distinguishing closely related taxa and contributes to the documentation and conservation of pteridophyte diversity in the Western Ghats.

DOI: https://doi.org/10.5281/zenodo.21990506

Unique Common Fixed Point of Multivalued Generalized ϕ-Weak Contractive Mappings with Integral Type Inequality

Authors: Dr. Manish Kumar Mishra

Abstract: The aim of this paper is to establish a unique Common Fixed Point of Multivalued Generalized ϕ-Weak Contractive Mappings in Integral Type Inequality. This result improves the result of Nadler and Daffer–Kaneko’s and references there in.

DOI: https://doi.org/10.5281/zenodo.21995390

An Explainable Stacking Ensemble Machine Learning Model For Denial-Of-Service Attack Detection

Authors: Halimah Muhammad Tukur, Jamilu Awwalu, Salim Ahmad, Ayuba John

Abstract: The evolving sophistication of cyber threats, particularly denial-of-service (DoS) and distributed denial-of-service (DDoS) attacks, poses significant challenges to the modern cybersecurity systems, availability, and reliability of digital infrastructures. These attacks often overwhelm network resources, leading to service disruptions and potential data breaches. Traditional intrusion detection systems (IDS) have been developed to mitigate such threats; however, they often struggle with scalability, detection accuracy, and interpretability, especially when processing large-scale network traffic. To address these limitations, this study proposes a stacked ensemble intrusion detection model integrating Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN) classifiers as base learners, with Logistic Regression (LR) serving as the meta-learner. The framework aims to enhance detection accuracy, interpretability, and robustness against DoS and DDoS attacks. Using the CIC-IDS2017 dataset, Mutual Information (MI) was employed for optimal feature selection, ensuring the most relevant attributes were retained for training. SHAP (Shapley Additive exPlanations) was applied to interpret the contribution of each selected feature to the model’s decision-making process. Experimental results demonstrate that the proposed ensemble achieves superior performance compared to individual classifiers and baseline models, with 98.83% accuracy, a marginal precision trade-off (96.37% vs. 96.50% baseline), 99.19% recall, 97.76% F1-score, and an AUC of 99.95%. Compared with the baseline [1], recall improved by 3.39% and F1-score by 1.57%. SHAP analysis identified flow-based and packet-based features as key contributors in distinguishing normal and malicious traffic. Overall, combining multiple learning algorithms within a stacked ensemble, along with MI feature selection and SHAP explainability, provides a robust, transparent, and effective solution for real-time intrusion detection.

DOI: https://doi.org/10.5281/zenodo.22011023

Generative AI Applications in Human Resource Management: Opportunities and Challenges

Authors: Assistant Professor Shahanas Kassim, Assistant Professor Soumya M K

Abstract: Generative Artificial Intelligence (GenAI) is revolutionizing the field of Human Resource Management (HRM) with the introduction of novel features in automation, personalization, and decision support in HR practices. The current paper provides an overview of the applications of GenAI in the context of HRM, analyzing their implications and challenges. With the help of literature review and a suggested methodological approach, we provide an in-depth overview of how GenAI technology is used to increase the efficiency of recruiting practices, personalize training programs, automate performance evaluation, and engage employees. The results of our quantitative research show that GenAI has the capability of reducing manual processing time by up to 90%, while at the same time increasing decision quality and mitigating the effect of biases in hiring decisions. Nonetheless, there are a number of important challenges associated with the use of GenAI technology, such as algorithmic bias, data privacy, ethical challenges, and workforce displacement. We suggest using an integrated approach to risk management in implementing GenAI in HRM.

DOI: https://doi.org/10.5281/zenodo.22011833

Strategic Workforce Planning in the Era of AI

Authors: Assistant Professor Dr. C. Sharmila, Assistant Professor D. Mercy Smilin

Abstract: Artificial Intelligence (AI) revolution is completely transforming the landscape of labor markets, generating the problem of workforce overprovision in traditional jobs and severe deficits of skills that are essential in the age of AI. This research proposes a quantitative approach for SWP that would solve the stated paradoxes using analytical and collaborative modeling of multiple scenarios of human-AI collaboration. Analyzing the survey results among 1,010 C-suite executives and the experience of firms that implement AI technologies, it becomes possible to prove that the firms implementing the approach of AI-driven workforce planning along with appropriate reskilling initiatives enjoy much higher levels of efficiency and resilience of their workforce. The suggested approach includes the DWAM model consisting of four interconnected modules, namely predictive demand sensing, skills gap analysis, reskilling optimization, and human-AI collaboration design. Simulations show that proper SWP allows reducing displacement rates by 40% and increasing talent retention by 35%.

DOI: https://doi.org/10.5281/zenodo.22012052

Data-Driven Financial Decision Making Using Machine Learning and Business Intelligence Tools

Authors: Research Scholar N. Rajarajeswari, Assistant Professor Ashwini DP

Abstract: Increase in financial data along with the advancements in computational intelligence have shifted the approach towards financial decision-making from an intuitive one to a data-driven one. This research paper aims to explore the usage of machine learning (ML) algorithms in combination with business intelligence (BI) techniques in order to improve the process of financial decision-making. An innovative framework involving the use of gradient boosting algorithms for prediction purposes and BI dashboards for visualization and interpretation is suggested. The framework includes such steps as data preparation, feature extraction, training and deploying the models within the BI framework. The results of analyzing the loan and investment datasets show that the proposed framework allows reaching a prediction accuracy of 89% and reduces decision-making time by 39%.

DOI: https://doi.org/10.5281/zenodo.22012319

Machine Learning-Based Student Performance Prediction and Early Warning System for Personalized Education Support

Authors: Atul Sharma, Aashish Kumar Tiwari

Abstract: Identifying students who are likely to struggle or disengage before the problem becomes visible in a final grade is one of the more practical uses of machine learning in education. This paper presents a student performance prediction and early warning system that consumes academic records and learning-management-system interaction logs, engineers a feature set describing both prior achievement and in-course behavior, and trains several classifiers — logistic regression, decision tree, random forest, support vector machine, and a small feed-forward artificial neural network — to flag students at risk of poor performance. The models were trained and evaluated on a held-out split of a simulated cohort constructed to resemble a real learning-management-system export, and compared using accuracy, F1-score, and ROC-AUC. The random forest classifier reached the highest accuracy at 91.6% and an AUC of 0.94, ahead of the neural network and support vector machine, while logistic regression trailed the ensemble methods but remained useful as an interpretable baseline. The resulting risk scores are surfaced through an advisor-facing dashboard rather than shown directly to students, so that a human can decide how and when to intervene. The paper describes the system architecture, the feature engineering and modeling pipeline, the evaluation results with supporting figures, and the practical and ethical constraints — class imbalance, explainability, and the risk of self-fulfilling labels — that should guide any real deployment.

DOI: https://doi.org/10.5281/zenodo.22025964

Financial Inclusion Through Digital Payment Innovations

Authors: Research Scholar N. Rajarajeswari, Assistant Professor Dr.Ramakrishna GN

Abstract: Various advancements in digital payment options have been developed as tools for transforming financially excluded communities across the globe. This work analyses the role those digital payments have played in extending financial services provision in developing countries, through a detailed evaluation of adoption factors, challenges faced and outcomes. Based on the empirical experience within the digital payment’s environment in India with respect to the use of Unified Payment Interface and RuPay cards, the study adopts an integrative quantitative and qualitative research strategy by analysing both transaction data and various stakeholder assessments. Results of the study show that there was almost 11 times increase in the adoption of digital payments in the period 2021-2025, especially as far as transaction volumes using UPI are concerned, accounting for almost 80% of total digital payments. Ease of use, trust, security, and digital financial literacy were found to be significant determinants of the use of these innovations.

DOI: https://doi.org/10.5281/zenodo.22030185

Cybersecurity And Ethical Hacking: A Proactive Defence Framework

Authors: Dangariya Dhrushal, Chandegra Rushil, Bhanderi Utsav, Prof. Sohil Parmar

Abstract: This paper examines the dynamic interaction between cybersecurity and ethical hacking, situating the latter as an indispensable, proactive measure of defence in a rapidly changing world of cyberspace. Cybersecurity, an interdisciplinary field with its roots in technology, is characterized by its fundamental principles of confidentiality, integrity, and availability. The paper explains the current taxonomy of cybersecurity disciplines, ranging from network to cloud security, and discusses varied adversary motivations, categorizing them as white, black, and gray hat hackers. Ethical hacking is also introduced as a systematic, five-stage process—reconnaissance, scanning, vulnerability assessment, exploitation, and reporting—modelled on actual attacks to determine vulnerabilities and remediate them. This process This process is not just a technical drill but a professional art guided by rigorous legal and ethical standards, underlining the paramountcy of permission and confidentiality. The document also reviews how new technologies such as AI and quantum computing are transforming threats and defences. In conclusion, the report determines that incorporating ethical hacking into a continuous strategy for managing cyber risk is critical for establishing organizational resilience, protecting digital assets, and achieving business continuity.

DOI: http://doi.org/10.5281/zenodo.22040505

Software Defect Prediction Using Machine Learning: A Comparative Analysis Of Classification Algorithms

Authors: Arjun Singh Tomar, Aashish Kumar Tiwari

Abstract: Identifying defect-prone modules before release is a central concern in software quality assurance, and machine-learning classifiers built on static code metrics have become a common way to prioritise testing effort. This paper presents a controlled comparison of six classification algorithms — Naive Bayes, Decision Tree, k-Nearest Neighbours, Logistic Regression, Support Vector Machine, and Random Forest — for software defect prediction on four datasets from the NASA/PROMISE repository (CM1, KC1, JM1, PC1). All models are trained and evaluated under an identical pipeline, including median imputation, feature standardisation, SMOTE-based resampling of the training folds, and stratified 10-fold cross-validation, and are compared using accuracy, precision, recall, F1-score, and ROC-AUC. Random Forest achieved the strongest overall performance (86.7% accuracy, 81.2% F1-score), followed by Support Vector Machine, while Naive Bayes and k-Nearest Neighbours lagged behind, particularly on precision. The results indicate that ensemble tree methods provide a robust default choice for defect prediction from static metrics, and that resampling strategy meaningfully narrows the performance gap between algorithms.

DOI: http://doi.org/10.5281/zenodo.22040921

Proximate Composition And Anti-Nutritional Properties Of Maize, Pigeon Pea, And Cooking Banana Flours For Composite Food Formulations

Authors: Agada Thankgod, Oloniyo Rebecca Olajumoke, Ugwuona Fabian Uchenna

Abstract: This study evaluated the nutritional and anti-nutritional properties of flours produced from maize, pigeon pea, and cooking banana to determine their suitability for use in composite food formulations. The flours were processed using standard methods and analyzed for proximate composition and selected anti-nutritional factors. The proximate composition (moisture, protein, fat, fibre, ash, and carbohydrate) and anti-nutritional factors (phytate, tannin, and haemagglutinin) were determined using standard analytical procedures, and the data obtained were subjected to statistical analysis to determine significant differences among the samples (p < 0.05). Significant differences (p < 0.05) were observed among the samples. Protein content ranged from 6.45% in maize flour to 7.45% in pigeon pea flour, while maize flour had the highest carbohydrate content (80.82%). Cooking banana flour contained the highest fat (4.12%) and fibre (2.24%) contents, whereas ash and moisture contents ranged from 2.35–2.74% and 5.75–7.23%, respectively. Phytate, tannin, and haemagglutinin contents ranged from 2.23–2.35 mg/100 g, 2.37–2.69 mg/100 g, and 0.74–0.84 Hiu/mg, respectively. Although pigeon pea flour recorded relatively higher anti-nutritional values, the levels observed were low and are expected to be further reduced during conventional food processing. The results demonstrate that each flour possesses unique nutritional attributes that can complement one another when combined. Maize provides a rich source of carbohydrates, pigeon pea contributes higher protein and mineral contents, while cooking banana enhances fibre content. These characteristics make the three flours suitable ingredients for developing nutritionally improved composite food products. The complementary nutritional characteristics of these flours suggest that their incorporation into composite formulations could enhance dietary quality, promote the utilization of locally available agricultural resources, and contribute to the development of affordable and nutritious food products for consumers. The findings provide useful baseline information for selecting appropriate raw materials and support the utilization of locally available crops in the development of value-added foods with improved nutritional quality.

DOI: http://doi.org/10.5281/zenodo.22042289

Pre-Conservative Optimisation In Musculoskeletal Physiotherapy: A Narrative Review Of Biomechanical, Psychological, And Systemic Readiness

Authors: Dr. Hardik Sheth (PT)

Abstract: Background: Musculoskeletal (MSK) conditions are a leading cause of global disability, with conservative management as the primary intervention. However, standard physiotherapy protocols often overlook baseline readiness, leading to poor adherence, high dropout rates, and suboptimal outcomes. While prehabilitation is established in surgical contexts, "pre-conservative optimisation “preparing patients prior to conservative care remains underexplored. Methods: This narrative review synthesised literature from PubMed, Embase, and CINAHL (January 2014–August 2026) using search terms related to musculoskeletal disorders, conservative management, patient readiness, and optimisation. Studies were thematically analysed to identify key factors influencing rehabilitation success prior to routine care. Results: Three critical dimensions of pre-conservative readiness were identified: 1) Biomechanical Readiness, establishing baseline load tolerance to prevent "boom-bust" cycles; 2) Psychological Readiness, leveraging Pain Neuroscience Education (PNE) and shared decision-making to reduce catastrophising and build self-efficacy; and 3) Systemic Readiness, optimising nutritional status for tissue healing and pain modulation. Integrating these dimensions creates a stable foundation for progressive therapeutic loading. Conclusion: Pre-conservative optimisation represents a vital paradigm shift in orthopaedic physiotherapy. Prioritising baseline readiness before standard protocols can enhance adherence, minimise symptom flares, and improve long-term outcomes. Future research should develop standardised assessment tools and evaluate the cost-effectiveness of this model.

DOI: http://doi.org/10.5281/zenodo.22043089

Fractional Moore–Gibson–Thompson Thermoelasticity in Functionally Graded Plates: A Semi-Analytical Differential Transform Approach

Authors: Dr. Sangita B. Pimpare

Abstract: This paper presents a semi-analytical Differential Transform Method (DTM) framework to analyze transient coupled thermoelastic fields in functionally graded (FG) rectangular plates under the fractional-order Moore–Gibson–Thompson (MGT) heat conduction model. The Caputo fractional operator models memory-dependent thermal transport, while power-law material distributions describe thickness-wise properties. Differential equations are converted into algebraic recurrences to evaluate non-Fourier responses without spatial discretization errors. Building on earlier analytical stress function approaches and reduced DTM schemes, this model extends MGT field analysis to functionally graded structures. Numerical results highlight how fractional memory, thermal relaxation, and material gradation affect thermal wave fronts and stress distributions.

DOI: https://doi.org/10.5281/zenodo.22056605

Applications of Parametric 2-Metric Spaces to Random Fixed Point Theorems and Pettis Integrability

Authors: Professor Dr. Dhirendra Kumar Singh, Virendra Kumar Saket

Abstract: The study of fixed point theory in generalised metric spaces has become an important topic in nonlinear analysis with applications to areas such as integral equations, optimisation and stochastic processes. We present new extensions of the framework of the parametric 2-metric spaces proposed by Cetkin (2019) in which we relate such a nonlinear geometry to measurable multifunctions and Pettis integrability in Banach spaces. First we recall the basic definitions and topological properties of parametric 2-metric spaces, and point out their role as a nonlinear generalisation of parametric metrics. Thus, we prove theorems on random fixed points and on the integrability of multifunctions in the setting of parametric 2-metric topology. The above results provide a unified framework connecting the theory of fixed points with the measurable selection theory and integrability conditions. We also discuss applications to integral inclusions and stochastic operator equations, showing that parametric 2-metric spaces can be a fruitful area for further research in functional analysis and stochastic settings.

Impact Of Polygamy On The Nutritional Status Of Children In Onna Local Government Area, Akwa Ibom State, Nigeria

Authors: Adasi, Owen Igakuboon

Abstract: Background: Family structure, particularly polygamy, remains a dominant socio-cultural feature in Sub-Saharan Africa. However, its influence on child nutrition and overall health outcomes in local settings like Onna Local Government Area (LGA) requires rigorous empirical evaluation. Objectives: To investigate the socio-economic status of polygamous parents, determine the impact of family size on child nutrition, and identify the health implications of malnutrition among children aged 0–18 years in Onna LGA. Methods: A descriptive cross-sectional survey was conducted using a multistage sampling technique to select 100 children (and their caregivers) across five administrative wards in Onna LGA. Data were collected using a structured, validated 15-item questionnaire, Impact of Polygamy on the Nutritional Status of Children Questionnaire (IPNSCQ), measured on a 4-point Likert scale (Test-retest reliability r = 0.81). Descriptive statistics using simple percentage analysis were employed to analyze the response patterns. Results: Of the 100 participants, 72% belonged to polygamous families. The socio-economic analysis revealed high rates of parental unemployment (36%) and self-employment/artisanal labor (54%). A significant majority of respondents agreed that low income in polygamous households hinders the provision of a balanced diet (81%) and leads to inter-sibling competition for food (88%). Large family size was strongly associated with inadequate feeding (81%) and routine meal-skipping (90%). Major health consequences identified included low immunity to infectious diseases (88%), emotional instability affecting academic focus (85%), higher risk of physical deformity (81%), and growth stunting (75%). Conclusion: Polygamous family structures compounded by low socio-economic status significantly impair child nutritional security and overall physical and cognitive health in Onna LGA. Targeted community health education, micro-economic policy support, and structural enforcement of family welfare frameworks are recommended.

Blockchain-Enabled Land Registry in Rural Bihar: An Analytical Study of Transparency, Fraud Prevention, Governance Efficiency, and Digital Inclusion

Authors: Pankaj Kumar, Professor Aparna, Dr. Abha Kumari

Abstract: This paper explores the deployment of a blockchain supported land-registry system in rural Bihar. In light of transparency, fraud mitigation, governance efficiency and digital inclusion, this paper refutes the common assumption that immutability of data results in accurate title. This study employs validated secondary data, from the years 2020-2025, such as the Bihar National Family Health Survey 2019-2021, various official sources of the Digital India Land Records Modernization Programme, Bihar land-service portals, and peer-reviewed literature on the intersection of blockchain and land governance. According to the National Family Health Survey (NFHS)-5, approximately 84 percent of surveyed households in Bihar were classified as rural, and a majority of the respondents, 79.4 percent of women and 56.4 percent of men, had never used the Internet. The widening of the access gap was examined in the context of the mobile phone ownership and usage, the financial inclusion of women, as well as the self-reported ownership of a house or land. The evidence-weighted readiness assessment determined that the level of digitization was relatively better, but the level of coordination of institutions, governance of cybersecurity, design of correction mechanisms, and design of user participation mechanisms were relatively poor. This paper proposes a permissioned industry consortium ledger, where sensitive data and documents remain off-chain, and the only data recorded on-chain are the hashes, identifiers, approvals, timestamps and version references of the land parcels. Smart contracts are used to manage the workflows from registration to mutation, but are not used to resolve the issues of contested titles, inheritance, or boundaries. This paper proposes an assisted-access model with a phased implementation approach, a multilingual interface, an appeal mechanism, and gender-disaggregated analysis and evaluation. Rather than fabricating field surveys and administrative performance data, this paper presents a complete primary data collection framework with a detailed statistical analysis plan for empirical assessment.

DOI: https://doi.org/10.5281/zenodo.22057349

Solid-State Cooling

Authors: Hassan S Alanzi

Abstract: Solid-state cooling technologies represent a transformative advancement in thermal management systems, particularly for energy-intensive industrial operations such as power plants, gas processing facilities, and utility infrastructures. Unlike conventional vapor-compression systems, solid-state cooling utilizes thermoelectric, magnetocaloric, electrocaloric, and electrocaloric effects to achieve heat transfer without refrigerants or mechanical compressors. This paper evaluates the technical principles of solid-state cooling and critically assesses its impact on energy efficiency, Energy Use Index (EUI), operational cost optimization, and utility infrastructure performance. The study finds that solid-state cooling can significantly enhance system reliability due to the absence of moving parts, reduce maintenance costs, and contribute to environmental sustainability by eliminating refrigerants. However, current limitations in coefficient of performance (COP) and scalability restrict widespread adoption in large-scale industrial applications. Despite these constraints, emerging innovations in materials science and control algorithms indicate strong potential for integration into decentralized cooling, electronics cooling, and hybrid plant systems. This paper concludes that solid-state cooling will play a strategic role in next-generation energy plant design, particularly in high-temperature regions such as the Middle East, Africa, Arizona state, Nevada deserts …where cooling loads dominate energy consumption. Future research should focus on improving efficiency metrics and integrating solid-state cooling within smart grid and digital plant ecosystems.

DOI: http://doi.org/10.5281/zenodo.22076425

Application of STATCOM for Voltage Stabilization in Long-Distance Transmission Systems: Optimized Scope, Safety and Cost

Authors: Hassan S Alanzi

Abstract: As global demand for reliable long-distance power transmission grows, particularly in the integration of renewable energy and the connection of industrial load centers, the need for advanced voltage control technologies has intensified. This paper explores the deployment of Static Synchronous Compensators (STATCOMs) within the context of project management for large-scale transmission projects. It examines the integration of STATCOM technology in the engineering, procurement, installation, and commissioning phases, with a focus on scope, schedule optimization, cost efficiency, safety enhancement, and alignment with international standards (NEC/IEC). The paper assesses key aspects of project governance, emphasizing the strategic planning and risk management considerations crucial for the successful execution of STATCOM installations.

DOI: https://doi.org/10.5281/zenodo.22076496

A Study of Yogic Therapeutic Intervention on Incontinence of Urine

Authors: Shridhar Ghodake

Abstract: Urinary incontinence (UI) is a common health condition affecting individuals across different age groups, leading to physical discomfort, psychological distress, and reduced quality of life. While conventional treatments such as medication and surgery are available, they may not always provide satisfactory results and can have side effects. The present study aims to explore the role of yogic practices in the management of urinary incontinence using a descriptive research approach. A total of 30 participants with symptoms of urinary incontinence were selected through purposive sampling. A structured yoga intervention consisting of asanas, pranayama, mudras, bandhas, and relaxation techniques was administered for 12 weeks. Observational data were collected using a structured questionnaire focusing on frequency, severity, and impact on daily activities. The findings indicate a reduction in urinary leakage, improved bladder control, and enhancement in quality of life among participants. The study suggests that yoga may serve as an effective complementary therapy for urinary incontinence.

Vehicle-Specific Underbody Collision Prediction Using Vision, Vehicle Dynamics, and Physics-Guided Machine Learning

Authors: Dr. Pankaj Malik, Yash Upadhyay, Swaraj Singh Rajput, Anuj Patidar, Taher Fakhri

Abstract: Underbody collision is a vehicle-specific safety problem encountered while traversing potholes, speed breakers, road humps, deep depressions, rocks, and uneven or off-road surfaces. Conventional vision-based advanced driver assistance systems (ADAS) can detect road anomalies, but anomaly detection alone does not determine whether a particular vehicle will experience underbody contact. This paper presents an Intelligent Vehicle-Specific Underbody Collision Prediction (IV-UCP) framework that integrates forward-camera road-obstacle detection, road-geometry estimation, vehicle geometry, vehicle dynamics, and a physics-guided minimum-clearance model. The framework represents road geometry, vehicle characteristics, and dynamic state in a multimodal feature vector and uses machine learning to estimate collision probability and risk level. A conservative safe-speed module, cross-vehicle evaluation protocol, ablation analysis, explainable AI, and real-time deployment evaluation are also incorporated. The physical criterion C_min = min_x[Z_u(x)-Z_r(x)] is used as an interpretable clearance feature, while uncertainty and ground-truth quality are treated as important experimental considerations. The numerical values currently retained in the manuscript are explicitly identified as tentative or target values where verified experimental measurements are not available; they must be replaced by measured results before making empirical performance claims. The framework therefore provides a structured path from road-anomaly recognition to vehicle-specific underbody collision-risk prediction and driver assistance.

DOI: https://doi.org/10.5281/zenodo.22094438

Crosswind Effects on the Aerodynamic Characteristics of the L-39NG During the Flare Phase of Landing Using ANSYS CFX

Authors: Trung Duc Pham

Abstract: In aviation operations, the landing phase, particularly the flare stage, plays a critical role, directly affecting flight training quality and flight safety. This paper investigates the influence of crosswind on the aerodynamic characteristics of the L-39NG aircraft using numerical simulation with Ansys software. The results show that crosswind increases drag, reduces lift, and generates lateral force, thereby affecting stability and controllability during landing. The study provides a scientific basis for improving operational efficiency and enhancing aircraft safety.

DOI: https://doi.org/10.5281/zenodo.22094684

School Scholars Exam Grade Prediction by Enhancing Privacy and Trained Model

Authors: Research Scholar Satish Kumar, Associate Professor Dr Surendra Singh Vishwakarma

Abstract: The growing availability of educational data has created new opportunities for predicting student academic performance, although privacy protection and appropriate feature selection remain significant challenges. This paper proposes a School Scholar Grade Prediction by Neural Network (SSGPNN) model for privacy-aware student grade prediction. Initially, distributed educational datasets are cleaned by removing sensitive information and transforming categorical attributes into suitable numerical representations. A Genetic Algorithm (GA) is employed to select an optimized set of relevant features and eliminate less informative attributes. The selected features are subsequently used to train a neural network for grade prediction. The proposed model is evaluated at dataset percentages of 12%, 25%, 50%, and 75% using precision, recall, F-measure, and accuracy. Experimental results show that SSGPNN outperforms K-PPD-ERT and achieves maximum precision, recall, F-measure, and accuracy values.

DOI: https://doi.org/10.5281/zenodo.22096044