Tianying Algorithm-Optimized XGBoost for Network Intrusion Detection
CAI Jiyong
Abstract：In response to the limitations of existing network intrusion detection algorithms,such as low accuracy and excessively long detection times,an XGBoost algorithm optimized using the Tianying algorithm has been designed.First,the data set entering the local network is preprocessed to improve its dimensionality.Then,the tree structure of XGBoost is established to split the objective function synchronously and evaluate the gap between the expected value and the output value.Finally,the Tianying optimization algorithm is introduced to dynamically select the key parameter values at different iteration stages.A weighted histogram is then used to accurately select the splitting points,controlling the scale of the tree structure and reducing the model's complexity.Experimental results demonstrate that the proposed optimization algorithm exhibits strong iterative efficiency.The accuracy rates of the training and test sets are 99.3％and 99.4％,respectively.It also has certain advantages in terms of controlling the detection time.Therefore,it can be concluded that the Tianying-optimized XGBoost algorithm is applicable to intrusion detection.
Keywords：XGBoost algorithmTianying optimization algorithmenhanced histogramiterative efficiency
Publication Date：2025-12-30
Online Publishing Date：2026-01-17(First online date of this platform, not the publication date of the document)
Pages：8( 42-49 )