International Research Journal of Education and Technology

The rapid growth of electronic commerce and digital banking has significantly increased the usage of credit cards for financial transactions. However, this expansion has also led to a substantial rise in fraudulent activities, causing serious financial losses and security concerns for both customers and financial institutions. Traditional fraud detection systems primarily rely on rule-based mechanisms or customer complaints, which often detect fraud only after the damage has occurred.

This paper proposes a Credit Card Fraud Detection System using a Hidden Markov Model (HMM) to identify fraudulent transactions based on cardholder spending behavior. The system models transaction sequences as probabilistic states and analyzes deviations from established behavioral patterns. Historical transaction data is used to train the HMM, enabling the system to evaluate the likelihood of new transactions in real time. Transactions with significantly low probability scores are flagged as suspicious and subjected to additional verification procedures. Experimental evaluation demonstrates that the proposed approach achieves approximately 80% detection accuracy while maintaining a reduced false positive rate.

Keywords: Credit Card Fraud Detection, Hidden Markov Model (HMM), Anomaly Detection, Behavioral Analysis, Machine Learning, Transaction Monitoring, Financial Security, Fraud Prevention, Real-Time Detection, Spending Pattern Analysis.