Fraud Detection in Banking Data by Machine Learning Techniques

More and more people are using credit cards to make purchases online, which has led to a surge in the number of financial transactions processed each day. Financial institutions must also bear high transaction costs due to the dramatic rise in fraud. Consequently, the art of fraud detection has emerged as a captivating field of study. In this research, we take into account the possibility of using hyperparameters for class weight adjustment to regulate the relative importance of valid and fraudulent transactions.Specifically, we optimize the hyperparameters using Bayesian optimization, which preservespractical concerns like imbalanced data. In order to address imbalanced data, we provide eight-tuning as a per-process.