A Performance Analysis Approach for Network Intrusion Detection Algorithms

With the development of mobile Internet and cloud computing, the amount of network traffic has been significantly increased. Security problems have drawn a lot of attention, while traditional methods are becoming increasingly unsuitable for it. In this paper, three machine learning algorithms are employed to detect network intrusion, including KNN, Random Forest, and Multilayer Perceptron. Performance evaluation and comparison between them are conducted, in terms of precision, recall, training time, etc. Simulation results on the NSL-KDD, a benchmark data set of network intrusion detection, show that the Random Forest algorithm exhibits higher detection accuracy and remarkably shorter training time.