International Journal of Science, Engineering and TechnologyAuthors: Dr.K. Geetha, Boomija.C. B, Dheepa Laksmi.A
Abstract: Within the contemporary digital landscape, individuals often encounter mental fatigue stemming from extended screen time, disrupted sleep cycles, and occupational stressors. The early identification of mental fatigue is crucial for sustaining productivity, emotional equilibrium, and overall health. Conventional wellness applications typically offer broad suggestions that do not account for individual variations in personality and lifestyle. This constraint diminishes the efficacy of these systems in mitigating mental fatigue[1]. This study introduces MindCare 360, an intelligent wellness framework that employs machine learning and explainable artificial intelligence (XAI) to identify mental fatigue and formulate personalized strategies for relief planner. The proposed system employs dual Random Forest classifiers to predict fatigue levels and personality types based on behavioral and lifestyle attributes such as sleep duration, stress levels, physical activity, social interaction, and screen time. Customized daily wellness strategies, segmented into morning, afternoon, evening, and night routines, are generated by integrating the outputs of these models through a rule-based decision-making process. Predictions, confusion matrices, and analyses of feature impact are presented via a Streamlit dashboard, designed with user experience as a priority. Furthermore, the Explainable AI component enhances transparency and builds user trust by elucidating the influence of various lifestyle factors like sleep hours ,sleep quality, and stress score based on their reactions to various situvations on fatigue predictions. Experimental assessments indicate that the proposed system accurately predicts fatigue levels while offering actionable and personalized relief suggestions. The MindCare 360 framework underscores the potential of integrating predictive analytics with personalized wellness planning to aid proactive mental health management.
DOI: https://doi.org/10.5281/zenodo.19111101
