Article’s

The rapid growth of digital banking and online payment systems has increased the volume of financial transactions worldwide. While these technologies provide convenience and efficiency, they have also created opportunities for fraudulent activities. Traditional fraud detection methods mainly depend on manual monitoring and predefined rules, which are often ineffective in identifying complex and emerging fraud patterns. Therefore, there is a need for intelligent systems that can analyze transaction behavior and detect suspicious activities in real time. This study presents a fraud detection system using Data Science and Machine Learning techniques to identify fraudulent financial transactions. The proposed system analyzes transaction attributes such as transaction amount, transaction type, sender balance, and receiver balance to predict the likelihood of fraud. A Random Forest classifier is employed to improve prediction accuracy, while a rule-based risk assessment mechanism helps categorize transactions into different risk levels. The system also provides real-time monitoring, analytics dashboards, alert generation, and report management features to support effective fraud investigation and decision-making.