Network Anomaly Detection Technology Based on Deep Learning
Adekunle Damilola Eunice, Qi Gao, Mengyuan Zhu, Zhuo Chen, Na Lv
Abstract
To improve the accuracy and real-time performance of anomaly detection models in complex network environments, a network anomaly detection model based on random forest and Deep Neural Networks (DNNs) is proposed. The feature importance is calculated based on the random forest algorithm. When detecting the anomaly traffic, we evaluate the performance of DNN from different layers. The NSL-KDD and USNW-NB15 datasets were used for training and testing. Experimental results show that a DNN of 4 layers has superior performance over all the other DNN layers algorithms in terms of stability and effectiveness, and feature selection method greatly reduces processing time by more than 50% with higher detection accuracy.