Fraud Detection in Banking Data by Machine Learning Technique

Aim:

          The aim is to leverage the power of machine learning to create efficient and accurate fraud detection systems that protect the interests of both financial institutions and their customers.

Abstract:

          As technology advanced and e-commerce services expanded, credit cards became one of the most popular payment methods, resulting in an increase in the volume of banking transactions. Furthermore, the significant increase in fraud requires high banking transaction costs. As a result, detecting fraudulent activities has become a fascinating topic. In this study, we consider the use of class weight-tuning hyper- parameters to control the weight of fraudulent and legitimate transactions. We use Bayesian optimization in particular to optimize the hyper parameters while preserving practical issues such as unbalanced data. We propose weight tuning as a pre-process for unbalanced data, as well as Cat Boost and XG Boost to improve the performance of the Light GBM method by accounting for the voting mechanism. Finally, in order to improve performance even further, we use deep learning to fine-tune the hyper parameters, particularly our proposed weight-tuning one. We perform some experiments on real-world data to test the proposed methods. To better cover unbalanced datasets, we use recall-precision metrics in addition to the standard ROC-AUC. Cat Boost, Light GBM, and XG Boost are evaluated separately using a 5-fold cross-validation method. Furthermore, the majority voting ensemble learning method is used to assess the performance of the combined algorithms. Light GBM and XG Boost achieve the best level criteria of ROC-AUC = 0.95, precision 0.79, recall 0.80, F1 score 0.79, and MCC 0.79, according to the results

Proposed System:

        Proposed approach to detecting credit card fraud using  a RFE (Recursive Feature Elimination) this feature selection method used to select the best feature for a dependent variable and also we used oversampling and then using a machine learning algorithm its give a best accuracy.