Credit card scam detection using machine learning

Finanical frauds have been a challenge to the corporate and government sector since a long time. Lately credit card based transactions have increased and so have the cases of fraudulent transactions using credit cards. These lead to massive financial losses in e-commerce and many other online payment methods. In this paper, we have implemented two machine learning technique – LOF(local outlier factor) and isolation forest to detect outliers and classify the transactions as fraudulent. fraudulent respectively. We obtained very good classification performance of 99% with the proposed method.