A very conventional fraud detection machine learning model based on various features such as new balance, old balance, amount transferred, etc. The model implemented are Random Forest, Decision Tree and XGBoost, it was found and explained that all except XGBoost can be considered for this purpose with certain degree of accuracy.