Research on Optimization of Abnormal Traffic Detection Model Based on Integrated Learning

In terms of the complexity and concealment of network attacks, an optimization research algorithm for an abnormal traffic detection model based on integrated learning is proposed. Firstly, the characteristics of network traffic data are extracted, and the data is preprocessed. Secondly, considering the problem of reducing the dimension of nonlinear noise, the kernel principal component analysis algorithm with standard deviation standardization is used to reduce the dimension of network abnormal traffic data. Finally, considering the limitations of a single model in determining classification problems, the method of an integrated model is adopted to classify abnormal traffic and improve the accuracy of classification. Simulation experiments show that the proposed algorithm can enhance the accuracy of abnormal traffic classification.