FCNN-SE: An Intrusion Detection Model Based on a Fusion CNN and Stacked Ensemble

As a security defense technique to protect networks from attacks, network intrusion detection model plays crucial role in the of computer systems and networks. Aiming at shortcomings complex feature extraction process insufficient information existing models, an named FCNN-SE, which uses fusion convolutional neural (FCNN) for stacked ensemble (SE) classification, is proposed this paper. The mainly includes two parts, classification. Multi-dimensional features traffic data are first extracted using different dimensions then fused into dataset. heterogeneous base learners combined used as classifier, obtained dataset fed classifier final comprehensive performance verified through experiments, experimental results evaluated evaluation method based on radar chart method. comparison NSL-KDD show that FCNN-SE has highest overall among all compared more balanced than other models.