Detection of Frauds for Debit Card Transactions at Automated Teller Machine in Indonesia Using Neural Network

Fraud detection in an online banking transactions such as in Automated Teller Machine (ATM) is one of the important strategy implemented by banks to protect customer's account. Fraud detection requires a lot of investments, complex algorithms, training and testing. Fraud destroy the reputation of banks, loss of financials, and loss of the country's finances. This research is conducted in order to propose a model using neural network and data mining to detect fraud in debit card transaction. Neural network can be used as a benchmark to develop a logistic regression model in data mining. The evaluation of performance classifier (accuracy, sensitivity and specificity) in this research showed that the proposed model predicted class label tuple correct (76.3%). The result supports fraud analysis in debit card transaction in ATM. In conclusion the results show that the model has a good performance in detecting fraud in a debit card transaction.