The Financial Fraud Detection Using A Value-At Risk With Machine Learning in Skewed Data

Financial fraud poses a significant threat to global markets, causing substantial economic losses and undermining investor confidence. Traditional fraud detection methods struggle with highly imbalanced datasets, where fraudulent transactions are rare compared to legitimate ones. This paper proposes a novel approach that integrates Value-at-Risk (VaR) with machine learning techniques to enhance fraud detection in skewed financial data. The system employs advanced machine learning algorithms, including Random Forest, Gradient Boosting, and Deep Neural Networks (DNN), optimized with synthetic data augmentation techniques such as SMOTE (Synthetic Minority Over-sampling Technique) and cost-sensitive learning. Experimental results on real-world financial datasets demonstrate that our approach improves fraud detection accuracy, reduces false positives, and provides a risk-based assessment of suspicious transactions.