Botnet detection method based on multi-mode stacked automatic encoder

The invention discloses a botnet detection method based on a multi-mode stacked automatic encoder. The method comprises the following steps: acquiring an executable file of an application program; respectively carrying out dynamic analysis and static analysis on a data set containing a benign program and a zombie program, and extracting dynamic features based on a flow and static features based on a printable character string information graph; pre-training two stacked automatic encoders, respectively encoding the stream-based features and the graph-based features, and extracting deep features; fusing the dynamic features and the static features based on a multi-modal automatic encoder; performing fine adjustment on the multi-mode stacked automatic encoder model; and taking an encoder of the trained multi-mode stacked automatic encoder model as a feature extractor, and taking the output of the shared hidden layer as the input of a softmax layer to carry out zombie program detection. According to the method, t