Improve the Abnormal Detection Method of Power Communication Data Traffic Based on LightGBM
Addressing the issue of variable attack modes and insufficient detection generalization ability in power communication networks.the abnormal detection method of power communication data traffic based on improved Light Gradient Boosting Machine(LightGBM)is proposed.Combined with Edited Nearest Neighbor(ENN)algorithm,wavelet packet decomposition technology and information entropy analysis method,extract the power communication data traffic abnormal features,introduce the histogram algorithm and with depth limit Leaf-wise growth strategy,improve the LightGBM algorithm for power communication data traffic anomaly detection model,find the optimal super parameter configuration in the model,improve the accuracy and efficiency of power communication data traffic abnormal detection.The experimental results show that the design method can significantly improve the gray detection rate and improve the detection effect,so as to effectively deal with the changing network threat.
万方数据
维普
