Fraudulent Transactions Prediction Using Deep Neural Network

Today, data is increasingly easily accessible, with corporations storing information with high volume, variety, speed, and value. This data is derived from several sources, including social media and user purchase transactions. Payment transactions have swiftly proliferated in their various forms,including cash-in, cash-out, debit, payment, and transfer. According to this, one of the most important dangers to online security nowadays is fraudulent transactions. Suspicious trans-action monitoring should therefore be an essential component of any payment system. Fraud transactions can be detected by evaluating consumer habits from past transaction data. In this paper we are proposed a Neural Network model to detect different types of fraudulent transactions using a recent data set to cope the weakness of the previous works. AUC of 99% has been obtained by using the proposed model as observed in the experimental results.