Abstract: The project's goal is to use machine learning models to predict fraudulent credit card transactions. From both the bank's and the customers' perspectives, this is essential. The banks are unable to bear the loss of consumer funds to dishonest individuals. Since the bank is in charge of the fraudulent transactions, every fraud results in a loss for the bank. This document can be used as a template to type your own text into or as a set of instructions. When developing the model, we must address the imbalance in the data and test different algorithms until we find the optimal model.
When developing the model, we must address the imbalance in the data and test different algorithms until we find the optimal model. Examining the valuable transaction and comparing it to a fresh, current transaction will reveal whether not the transaction is fraudulent. The purpose of this paper is to provide a comprehensive review of various techniques for detecting fraud. We suggested a system that uses a decision tree, random forest, and logistic regression to identify fraud in the processing of credit card transactions. After then, we use XGBoost to fit the dataset and categorize the transactions as either fraudulent or not. Because we were working with a heavily skewed dataset, we used the F1 and ROC AUC scores to evaluate our model's performance.

Keywords: Fraud Detection Techniques, Logistic Regression, Random Forest, Decision Tree, XGBoost Classifier.

Works Cited:

Dr. Dinesh D. Patil, Dr. Priti Subramanium, Pooja Balu Wankhede " Using a Machine Learning Model to Predict Fraudulent Credit Card Transactions ", IARJSET International Advanced Research Journal in Science, Engineering and Technology, vol. 10, no. 11, pp. 70-75, 2023. Crossref https://doi.org/10.17148/IARJSET.2023.101110

| DOI: 10.17148/IARJSET.2023.101110