Risk Assessment Model of Information Base Based on Machine Learning in Big Data Environment

How to effectively assess risk in information databases has emerged as a prominent research topic as big data technologies evolve and quantum computing rises. This study used quantum neural networks as the core tool and combined them with self-organizing feature map neural networks to jointly construct an information-based risk assessment model. Results showed that in convergence comparison when the number of iterations was 42, the research method had a maximum fitness value of 99.99 on the training set. As the sample size increased, the gap between the research method and the expected value gradually decreased, reaching a high fitting degree of 0.972. The fitting degrees of the three models combined with adaptive quantum neural network and chaotic synchronization algorithm, short-term memory convolutional neural network, and IK-means were 0.685, 0.712, and 0.776, respectively. When the accuracy of all methods was 0.800, the recall rates of the research method, the combination of adaptive quantum neural network and chaotic synchronization algorithm, and the long and short-term memory convolutional neural network and IK-means were 0.901, 0.849, 0.799, and 0.687, respectively. During the application test, the proposed method achieved over 92% in all three tests, with a maximum of 97% and higher accuracy. The above results indicate that the research method has higher computational efficiency and prediction accuracy compared to traditional methods when dealing with large-scale datasets. It can effectively assess the risks of the information base. © (2024), (International Journal of Network Security). All rights reserved.