International Journal of Science, Engineering and TechnologyAuthors: Priyesh Mahajan, Nitin Namdev
Abstract: Credit card fraud is a growing problem in today’s digital world. It causes heavy financial losses for companies and puts consumer security at risk. Traditional rule-based systems help to some extent, but they often fail against advanced fraud tricks.Machine learning (ML), especially anomaly detection, offers a better solution. It can spot unusual patterns in large datasets, making fraud detection faster and smarter. Researchers are exploring many methods, such as autoencoders, hybrid models, and new feature selection techniques.Still, there are challenges. Fraud data is highly imbalanced, and real-time detection is difficult to achieve. Even so, ML-based anomaly detection shows great promise. With continuous improvements, it could make financial transactions much safer in the near future
