Research on Intrusion Detection Model Based on PCA-SVM
Lei Pan, Tianyi Zheng, Rongyao Fu · 2022 IEEE 4th International Conference on Civil Aviation Safety and Information Technology (ICCASIT) · 2022
With the development of machine learning technology, the application of intrusion detection models with statistical machine learning as its core in network security issues shows important research significance. To aim at the problem that the accuracy and efficiency of basic SVM algorithm in detecting NSL-KDD data sets are not ideal, this paper designs a PCA-SVM phased intrusion detection model. Firstly, the dimension of the data set is reduced based on principal component analysis, and then the model parameters are obtained by training SVM with the NSL-KDD test data set. Finally, the experiment is carried out on the NSL-KDD test set. Compared with the SVM single model, when the penalty coefficient is 2.0, the accuracy of the phased model is increased from 65.6% to 86.7%, and the detection time is reduced by 27.0%.