Fraud Detection in Banking Industry using Enhanced Intellectual Framework with Optimization Strategy

The financial industry has become essential in the modern era, as nearly all individuals have to interact with a bank, neither physically nor digitally. Online fraud happens when credit card information gets shared with an unauthorized individual through the Internet or mobile. Both consumers and financial institutions might face multiple problems because of criminals and scammers while performing transactions and the probability of getting trapped is extremely high. Therefore, identifying such kinds of fraud is crucial. In this paper, a unique deep learning-aided technique is suggested to detect fraud in the banking sector to prevent financial loss. At first, the data is obtained for processing are collected from online resources. From the accumulated data, the essential features are extracted using an autoencoder. The retrieved features from the autoencoder are passed to the Adaptive Gated Recurrent Unit (AGRU) for detecting fraudulent activity in the banking sector. To attain better performance in fraud detection, the parameters in the AGRU are tuned via the Election Optimizer Algorithm (EOA). The recommended system overcomes fraudulent analysis in banking sector and it enhances the confidence of the user. Finally, the effectiveness of the designed framework is analyzed over the diverse methods to confirm its efficiency.