Credit Card Fraud Detection Using Machine Learning

Abstract

Frauds are the  biggest threats in our day-to-day life is Fraud. Nowadays frauds are spreading all over the world and due to this financial loses are increasingly high in all sectors. Malicious behavior may be a broad term together with delinquency, fraud, intrusion, and account defaulting. Many banks find it difficult to detect the fraud in credit card system. The fraud detection performance in credit card transactions is highly affected by the sampling approach on data-set, variable selection and detection techniques. In order to resolve this problem this project comes with the fraud detection techniques that could bring a solution to this problem with a predictive analysis. A predictive model is being developed for credit card fraud detection using Spyder IDE and various machine learning algorithms are applied for the best identification of the frauds with the best accuracy for the dataset taken. The methods used in fraud detection evaluates a new hybrid approach to identify fraud detection. The credit card fraud detection is employed using machine learning algorithm namely Decision Tree, logistic regression, Random Forest and Support Vector Machine. To make the learning process efficient, we used Principal component for feature selection and a comparison is made on different machine learning algorithms with different parameters for better accuracy.