Aiming at the problems of incomplete feature information extraction and low accuracy of multi-classification detection in current network intrusion detection methods,a hybrid network intrusion detection method based on CNN-BLSTM-XGB was pro-posed.A network structure CNN-BLSTM based on the combination of convolutional neural network and bidirectional long short-term memory network was established to extract the spatial and the temporal features of network intrusion data.The Concate-nate layer in the Keras sequential model was used to fuse the two features.The extreme gradient boosting was used to replace the traditional complete connection layer to obtain the feature information from the input layer to the fusion layer for classification.Experimental results on NSL-KDD and CICIDS2017 datasets show that the multi-classification detection accuracies of proposed method can achieve 99.72％and 99.87％respectively,which are higher compared with that of existing methods.
维普
万方数据