Various Credit Card Fraud Detection Techniques based on Machine Learning

Abstract

Click Here:Access Full Text

Use of mobile devices facilitates more people to do online shopping using credit cards. As a result, Internet shopping has become a popular method of making daily purchases. Online shopping offers benefits like efficiency, convenience and greater selection as well as better pricing. With the number of transactions by credit cards are increasing rapidly transaction fraud are also increasing. A fraud activity results in financial loss to the individuals. Therefore financial institutions provide more value and demand for fraud detection applications. We need to also make our systems learn from the past submitted frauds and make them fit for adapting to future new methods of frauds. This paper shares the concept of frauds related to credit cards and their various types. We have considered various techniques available for a fraud detection system such as, Hidden Markov Model (HMM), Bayesian Network, Hybrid Support Vector Machine (HSVM), K-Nearest Neighbor (KNN), Naïve Bayes, Logic regression, Decision Tree and a feedback mechanism. This paper covers existing and proposed models for credit card fraud detection and focus to find the best method by comparing different techniques on the basis of quantitative estimations like accuracy, detection rate and false alarm rate.

Keywords: Bayesian network, Hybrid support vector machine, K-Nearest neighbor, Logic regression, Naïve Bayes.

View PDF