Published Paper Details:

As digital banking becomes increasingly central to financial activity, the risk of fraud has grown significantly. Traditional fraud detection methods, which often rely on static rules, are proving insufficient against the sophisticated strategies of modern cybercriminals. This research focuses on utilizing machine learning (ML) approaches to enhance the detection of fraudulent behaviour in online banking systems. By examining extensive transaction datasets, ML models can learn to recognize unusual patterns that may signal fraud, allowing for timely and accurate detection. The system developed in this study applies supervised learning techniques such as logistic regression, decision trees, and random forest classifiers trained on labelled data to distinguish between legitimate and suspicious transactions. Through careful feature extraction and data pre-processing, the models achieve improved precision and recall. Findings indicate that machine learning offers a dynamic and efficient framework for combating fraud, ultimately strengthening the safety of online banking and fostering greater user confidence.