Development of machine learning based fraud detection models for credit cards

In today\\\'s global world, technology is rapidly developing and this can cause more risks, especially in sectors such as banking. Fraudsters create security vulnerabilities with many new techniques. Various approaches have emerged to prevent these vulnerabilities, but these approaches are generally inadequate due to reasons such as high data volume, multiple institutions, channels (mobile applications, websites, call centers) and fraudulent activities between locations. In this context, machine learning-based systems gain importance due to their dynamic structure. In this study, it is aimed to develop a model that provides fraudulent transaction detection using the Random Forest (RF) classifier. Docker and Kubernetes have been used for model distribution in the study. The performance of the developed model has been evaluated with Accuracy, Precision, Recall and F1 Score. With the developed fraud detection model, an Accuracy value of 0.771 has been achieved.