International Journal of Smart Business and Technology Vol.9, No.2 (2021), pp.33-48 http://dx.doi.org/10.21742/ijsbt.2021.9.2.03 Print ISSN: 2288-8969, eISSN: 2207-516X IJSBT Copyright ⓒ 2021 Global Vision Press (GV Press) Ensemble Techniques for Credit Card Fraud Detection Satya Dileep Penmetsa 1 and Sabah Mohammed 2* 1,2* Computer Science Department, Lakehead University, Canada 1 penmetsas@lakeheadu.ca, 2* sabah.mohammed@lakeheadu.ca Abstract Credit card fraud is a problem that has grown by great danger and has a huge impact on the financial sector. The challenges of credit card fraud are the availability of public data, high imbalance in data, and volatility of the fraud nature. Over the years ensemble learning has gained more importance and proved to give better performance. Here we try to do a comparative study of various ensemble approaches using various learning algorithms on the credit card fraud data and to understand multiple models based on various evaluation and performance metrics using the SMOTE balancing technique. Keywords: Credit card fraud, Machine learning, Ensemble learning, Oversampling, SMOTE 1. Introduction Over the years, due to the rise of e-commerce, the use of debit cards for purchases has increased drastically. The present unprecedented just added more and has increases the use of credit cards manifold. Credit card payments have become one of the popular methods of purchase and have revolutionized the way of payments. Financial institutions issue these credit cards for the customers to use, the customers can use these cards for purchases and doing credit transactions by using the details imprinted upon the cards. When the information of your credit card is used to make purchases without your knowledge that implies there is a fraudulent transaction which can adversely cost you money and can also affect your credit score. Some individuals try to exploit this information making identity thefts and cause of huge losses to credit card owners by doing fraudulent transactions and this problem is something to be addressed with the increased e-commerce and the ease of making payments and transactions, it is now important than ever before for establish proper fraud detection and fraud prevention techniques. While fraud prevention works by setting some thresholds and security methods to prevent fraud, fraud detection is totally on understanding the patterns of fraud transactions by Machine Learning and Deep Learning techniques to check for the possibility of a fraudulent transaction. As the methods and strategies used by the fraudsters change constantly there is a high and constant need for ML techniques in the sector. And over the years these machine learning techniques have been widely used in fraud detection and achieved favorable performances. Article history: Received (April 14, 2021), Review Result (May 16, 2021), Accepted (August 30, 2021)