Fraud detection has appertained to many industries such as banking, retails, financial services, healthcare, etc. As we know, fraud detection is a set of campaigns undertaken to avert the acquisition of illegal means to obtain money or property under false pretense. With an unlimited and growing number of ways fraudsters commit fraud crimes, detecting online fraud was so tricky to achieve. This research work aims to examine feasible ways to identify credit card fraudulent activities that negatively impact financial institutes. In the United States, an average of U.S consumers lost a median of $429 from credit card fraud in 2017, according to “CPO magazine. Almost 79% of consumers who experienced credit card fraud did not suffer any financial impact whatsoever” . One of the questions is, who is paying for these losses if not the consumers? The answer to this question is the financial institutions. According to the Federal Trade Commission report, credit card theft has increased by 44.6% from 2019 to 2020, and the amount of money lost to credit card fraud in the year 2020 is about 149 million in total loss. Without any delay, financial institutes should implement technology safeguards and cybersecurity to decrease the impact of credit card fraud activities. To compare our chosen machine learning algorithms with machine learning techniques that already exist, we carried out a comparative analysis and we were able to determine which algorithm can best predict fraudulent transactions by recognizing a pattern that is different from others. We trained our algorithms over two sampling methods (undersampling and oversampling) of the credit card fraud dataset and, the best algorithm is drawn to predict frauds. AUC score and other metrics was used to compare and contrast the results of our algorithms. The following results are concluded based on our study:
1. Our study proposed algorithms such as Random Forest, Decision Trees and Xgboost, K-Means, Logistic Regression and Neural Network have performed better than other machine learning algorithms researchers have used in previous studies to predict credit card frauds.
2. Our ensemble tree algorithms such as Random Forest, Decision Trees and Xgboost came out to be the best model that can predict credit card fraud with AUC score of 1.00%, 0.99% and 0.99% respectively.
3. The best algorithm for this study shows a lot of improvements with the oversampling dataset with overall performance of 1.00% AUC score.
Date of Degree
Master of Science (MS)
Program of Study
Computer Information Science
Science, Engineering and Technology
Ayorinde, K. (2021). A methodology for detecting credit card fraud [Master’s thesis, Minnesota State University, Mankato]. Cornerstone: A Collection of Scholarly and Creative Works for Minnesota State University, Mankato. https://cornerstone.lib.mnsu.edu/etds/1168
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-Share Alike 4.0 International License.