Commercial Bank IT Risk Evaluation Model Based on GA-BP Neural Network

To assess IT risk in commercial banks and reduce the probability of incidents, we constructed a set of IT risk evaluation indices based on determining the set of IT risk factors. To improve the accuracy of the evaluation model, the Genetic Algorithm (GA) was used to adjust the weights and thresholds of the BP neural network, thereby establishing a GABP-based IT risk evaluation model. By collecting data from a commercial bank for model testing, the IT risk evaluation was successfully implemented, and the weights of risk evaluation indicators were calculated. The correlation coefficient R values of the model’s training set, test set, validation set, and full sample set were greater than 0.95, demonstrating excellent predictive performance and effective IT risk evaluation. Based on the evaluation indicators, the weights of the indicators, and risk levels, commercial banks can develop corresponding targeted preventive and control measures. This comprehensive approach provides commercial banks with more reliable IT risk management and decision support.