Abstract : To make life better, many mechanisms in modern environment are carried out via the Internet. The economy is expanding yet on the other side, there is a lot of illegal and unauthorised activity carried throughout the country that is seriously hampering that progress. Scam instances, which mislead individuals while also causing economic losses, are just one of them. In realistic conditions, fraud involving credit cards surveillance is the main emphasis of this research. Contrary to earlier eras, the number of credit card scammers is drastically increasing right now. Criminals use various forms of innovation, fake documents, and deception to con others and take their cash. Therefore, it is extremely crucial to discover a solution to these frauds. As technology advances, it becomes harder to keep up with the behaviour and trends of illegal activities. Ai technology, machine learning, as well as other relevant data technology fields have advanced to the point that it is currently feasible to expedite this process and reduce the volume of labour-intensive effort needed in recognizing credit card scams. The user-submitted utilization of credit cards database might be collected initially, then using machine learning approach; it would be split into databases for testing and training purposes. This methodical technique could be utilized by researchers once they have evaluated both the larger information collection and the user-provided available data collection. Enhance the accuracy of the outcome statistics after that. Depending on its exactness and precision, a technology's efficiency is assessed. The results show that XG-Boost and Random Forest techniques have the greatest performance.

Keyword : Credit card; XG-Boost, Fraud Detection, Machine Learning Techniques, Random Forest method.