Credit Card Fraud Detection using GSA

Abstract

Billions of dollars of loss are caused every year by fraudulent credit card transactions. The design of efficient fraud detection algorithms is the key for reducing these losses, and more and more algorithms rely on advanced machine learning techniques to assist fraud investigators. We found German credit card fraud detection database available publically which is having 1000 data points. This dataset is divided into 70/30 ratio for training and testing the neural network. The famous and efficient machine learning neural network algorithm is used to get a trained NN. This network is further updated for more classification accuracy using Gravitational Search Algorithm (GSA) which is an optimization algorithm. It tunes NN's weights and biases and check for the mean square error which is an evaluation parameter also in our work. Complete work is simulated in MATLAB R 2016a. Results are compared with previously used simulated annealing (SA) algorithm and proposed method is giving better results in term of area under curve (AUC) of ROC (receiver operating characteristics) and MSE.