Detection And Prevention Of Financial Fraud In Banking Transactions Using Neural Networks: A Case Of Cabs Bank

Author

Melissa Marimbi

Title

Detection And Prevention Of Financial Fraud In Banking Transactions Using Neural Networks: A Case Of Cabs Bank

Abstract

Financial fraud poses a significant threat to the stability and integrity of banking institutions worldwide. With the increasing volume and complexity of digital transactions, traditional rule-based fraud detection systems are often inadequate in identifying sophisticated fraud patterns. This study explores the application of neural networks to detect and prevent financial fraud in banking transactions at CABS Bank Zimbabwe. The research employs a mixed-methods approach, combining quantitative analysis of transaction data with qualitative insights from bank employees and customers. Transactional data from CABS Bank is used to train and evaluate a neural network model, focusing on features such as transaction amount, frequency, geographical location, and timing. The model's performance is compared with existing fraud detection methods, demonstrating significantly higher accuracy and reliability in detecting fraudulent transactions. Key findings reveal that neural networks can effectively identify complex fraud patterns and process transactions in real-time, providing immediate alerts for suspicious activities. The study also highlights the importance of feature engineering and continuous model updates to adapt to evolving fraud tactics. Interviews with bank employees underscore the need for integrating advanced fraud detection systems into existing workflows and ensuring staff training for optimal utilization. The implementation of neural network-based fraud detection at CABS Bank promises to enhance the bank’s security measures, reduce financial losses, and improve customer trust. This study concludes with practical recommendations for the bank, including adopting neural networks, enhancing data collection, and conducting regular employee training. Suggestions for future research include exploring advanced models, improving model interpretability, and examining the scalability of neural network systems. By leveraging the power of neural networks, CABS Bank Zimbabwe can significantly bolster its fraud detection capabilities, safeguarding financial assets and reinforcing customer confidence.

Date

June 2024

Publisher

BUSE

Keywords

Financial Fraud

Neural Networks

Fraud Detection,

Machine Learning

Supervisor

Mr B. Kusotera