International Research Journal of Education and Technology

As the number of companies and consumers using credit card information for financial transactions has grown, which has resulted in a substantial rise in fraud cases. This issue has been made worse by dealing with outliers, unbalanced, and noisy data. This work proposes the use of artificial intelligence for fraud detection. The proposed technique uses logistic regression as a method for creating the classifier in order to avoid transactions using credit cards fraud. Pre-processing ensures a high degree of precision in detection and is employed to control dirty data. The pre-processing stage of cleaning the data uses two distinct basic techniques: the mean-based methodology and the clustering-based strategies. The proposed classifier yields better results in terms of error rate, receptivity, and accuracy.

Keywords : Credit cards, machine learning, and artificial intelligence