Fraud Detection in Financial Transactions Using Machine Learning Techniques

Abstract—Through integrating data warehousing, data visualization, information retrieval, and stream processing analytics, this project seeks to create a strong fraud detection framework for financial institutions. Significant volumes of transactional data are efficiently processed and stored, allowing for rapid retrieval and in-depth analysis. To detect trends and irregularities, advanced machine learning models such as Random Forest, Decision Tree, Logistic Regression & Naïve Bayes are used. To evaluate these models, we use precision, recall, F1-score, and AUC-ROC. To help ensure practical deployment in financial environments, the system places a strong emphasis on scalability and real-time processing. This improves fraud prevention and strengthens confidence in economic systems. Keywords: K-Means, Financial Transactions, Random Forests, Decision Trees, Ensemble Learning, Logistic Regression, Fraud Detection, and Classification.