Network intrusion detection method, network intrusion detection model training method, network intrusion detection model training device and server

The invention provides a network intrusion detection method, a network intrusion detection model training method, a network intrusion detection model training device and a server. The network intrusion detection method comprises the steps of acquiring to-be-detected network traffic data; extracting feature data of the network traffic data through a pre-trained intrusion data detection model, and determining whether the network traffic data is intrusion data based on the feature data, wherein the intrusion data detection model is obtained by training a bidirectional long-short-term memory neural network, and the intrusion data detection model is further used for setting a preset weight value for the extracted feature data. An attention mechanism is added to the detection model, all featurescan be fused, the key features can be endowed with higher weights, then the intrusion data sequence information extraction capacity of the model is enhanced, and the intrusion detection accuracy is improved. 本发明提供了一种网络入侵检测方法