Research on intelligent detection of network attack based on XGBoost
Siqi Huang, Jiongqing Cao, Yang Yu, Xing Du, Jiankun Yuan, Kai Yu · Procedia Computer Science · 2025
At present, most of the detection methods for network attacks are based on traditional eigenvalue statistics or machine learning. Due to the lack of in-depth understanding of the data flow itself and the insufficiency of data feature extraction methods, the detection effect of network attacks is not good. To solve the above problems, this paper proposes an intelligent detection method of network attack based on XGBoost model. Firstly, multi-task learning technology is used for data preprocessing and feature extraction. Then XGBoost algorithm is used to train the detection model. Finally, experiments are performed on real data sets. And compared with decision tree algorithm and random tree algorithm, the experimental results show that this method can effectively detect network attack behavior and has higher accuracy than the other two methods.