Research Project: Comparison of Machine Learning and Deep Learning Techniques for Transaction Fraud Detection
Conducted a comparative experiment to train and evaluate multiple machine learning models using transaction data for fraud detection. The work involved data preprocessing, cleaning, feature engineering, and addressing class imbalance to prepare datasets for model training. Evaluated and compared the effectiveness of each model to recommend the optimal approach for AI-driven fraud detection. • Labeled transactions as fraudulent or non-fraudulent for supervised training. • Used common metrics such as Accuracy, Precision, Recall, F1-score, and ROC-AUC for evaluation. • Integrated AI training recommendations into a mobile banking context. • Explored both shallow and deep learning classification methods.