Credit Card Fraud Detection using Machine and Deep Learning Techniques
Shagun Sharma, Anjali Kataria, Jasminder Kaur Sandhu, K. R. Ramkumar
Abstract
With the increase in smart services, it became very easy to do online credit card transactions via IoT environment. It has facilitated people to do transactions under a short span of time. Along with the electronic facilities there are many issues and challenges, which take place during online transactions. These challenges could cause a huge loss in data and billions of dollars. In the year of 2017, it was identified that financial institute of North America has faced a loss of 3 billion dollars due to credit card scams. These scams could happen frequently and result worst in the area of finance. There are many technologies like Phishing or Trojan, which could be used by hackers/criminals to steal credit card holder’s personal information. Therefore, an effective method is needed to deal with the credit card challenges to solve the problem of personal information leakage. This, article discusses various Machine Learning classification algorithms for identification of credit card scams. Python is used to carry out the work whereas accuracy and precision metrics are used to evaluate the performance of each algorithm. The results demonstrate that the KNN classifier performs best with an optimal accuracy and precision of 98.33% and 0.5504 respectively. The results demonstrate that K-Nearest Neighbor outperforms as compared to other classifiers taken for the identification of credit card fraud in the article.