Fraud Detection using Recurrent Neural Networks for Digital Wallet Security

E-wallet use has exploded, bringing in a new era of convenient and secure digital transactions; nevertheless, it has also led to a surge in fraudulent transactions. Using Recurrent Neural Networks (RNNs), this research provides an experimental approach for detecting fraudulent use of e-wallets. To detect fraudulent behaviors in e-wallet transactions, we have developed Long-Short Term Memory (LSTM) architecture to make use of the sequential nature of transaction data. Using the United Payment dataset for training and testing, the proposed model successfully classifies fraudulent transactions while reducing the number of false positives. The findings underline the utility of the RNN-based strategy in strengthening the safety and reliability of e-wallet systems, lead off for more secure online financial dealings. This study adds to the growing body of literature on detecting fraudulent use of electronic wallets by developing a model that makes use of deep learning to enhance fraud detection effectiveness in the ever-changing domain of digital payments.