Research on Network Intrusion Detection Based on an Improved Deep Learning Method

Network intrusion detection is an important research direction in the field of network security. The traditional algorithm based on feature extraction and separation, which has problems low accuracy high false alarm rate. In order to improve detection, this paper proposes model deep asymmetric convolutional encoder Random Forest(RF). First, use DACAE extract features from preprocessed data, then random forest divide traffic data into normal abnormal classes, finally achieve purpose detection. It tested three public benchmark datasets NSL-KDD KDD99 datasets. experimental results show that rate improved method are better than comparative method.