Optimization of Algorithm for Network Traffic Anomaly Detection Using Convolutional Neural Networks (CNN)

With the rapid development of network technology, network security issues have become particularly prominent, especially in the context of big data and the Internet plus era, traditional network anomaly detection methods have been difficult to meet the needs of modern network security. This article studies a convolutional neural network model suitable for traffic data features, which can effectively extract spatiotemporal features from traffic data; an innovative data preprocessing technique is introduced to enhance the model’s generalization ability through data augmentation and feature standardization. The research results indicate that compared with traditional rule-based and machine learning-based detection methods, the method proposed in this study has significantly improved detection speed and accuracy. In the absence of any noise, the accuracy of the model reached 99.0%, demonstrating its excellent performance on pure datasets. This study not only optimized the application of CNN (Convolutional Neural Network) in network traffic anomaly detection, but also provided theoretical basis and experimental support for the development of future network security technology.