Published Paper Details:

Abstract- Credit card fraud is a serious risk to financial institutions and consumers, resulting in huge monetary losses globally. Rule-based fraud detection systems often fail to respond effectively to changing fraudulent patterns, making machine learning (ML) a viable alternative. The present work investigates the use of diverse ML algorithms such as Logistic Regression, Decision Trees, Random Forest, Support Vector Machines (SVM), and Neural Networks to identify fraudulent credit card transactions. The research utilizes a real-world dataset with highly imbalanced class distribution, which requires advanced data preprocessing techniques such as Synthetic Minority Over-sampling Technique (SMOTE) to improve model efficiency. Several evaluation metrics, such as Precision, Recall, F1-score, and the Area Under the Receiver Operating Characteristic (AUC-ROC) curve, are used to measure model efficacy. Experimental outcomes illustrate that ensemble learning methods, specifically Random Forest, perform better than other classifiers in terms of accuracy and fraud detection ability. The results reflect the strengths of ML- based fraud detection systems to detect fraudulent transactions with better precision and avoid false positives. Still, challenges such as interpretability of the model and limitations on real-time detection are topics of future work. This research presents useful findings towards improving fraud detection mechanisms and identifies future work on optimizing ML strategies in financial security.