SnapML library's Decision Tree classifier and SVM was used to train a model on a real dataset to identify fraudulent credit card transactions. The Decision Tree model resulted in ROC-AUC score = 0.92 and the SVM yielded ROC-AUC score = 0.93 and hinge loss = 0.15.
Multi-threaded CPU was implemented to reduce model training time.
The dataset includes information about transactions made by credit cards in September 2013 by European cardholders. The Kaggle dataset "Credit Card Fraud Detection" can be downloaded from the following link: Credit Card Fraud Detection