Abstract
Recently, network attacks launched by malicious attackers have seriously affected modern life and enterprise production, and these network attack samples have the characteristic of type imbalance, which undoubtedly increases the difficulty of intrusion detection. In response to this problem, it would naturally be very meaningful to design an intrusion detection system (IDS) to effectively and quickly identify and detect malicious behaviors. In our work, we have proposed a method for an IDS-combined incremental extreme learning machine (I-ELM) with an adaptive principal component (A-PCA). In this method, the relevant features of network traffic are adaptively selected, where the best detection accuracy can then be obtained by I-ELM. We have used the NSL-KDD standard dataset and UNSW-NB15 standard dataset to evaluate the performance of our proposed method. Through analysis of the experimental results, we can see that our proposed method has better computation capacity, stronger generalization ability, and higher accuracy.
Highlights
With the development of modern cyber-technologies, Internet technology has developed rapidly and we have entered an era of interconnection with everything
We demonstrate a method named incremental ELM (I-ELM), which is an improvement from ELM, and adaptive principal component analysis (A-Principal component analysis (PCA)), which is a combination of the adaptive control idea and PCA
Adaptive principal component analysis (A-PCA) is a method that combines the adaptive control theory with PCA, which selects the features after being decomposed by PCA by comparing the given performance indicators we set by automatically adjusting the step size α of r to compress the dataset according to the value of Acc and dimension after using PCA
Summary
With the development of modern cyber-technologies, Internet technology has developed rapidly and we have entered an era of interconnection with everything. IDS, proposed by Anderson [1], is a method/way to protect application systems from malicious attacks, which is considered as the second defending line. The updated hacker technology and powerful attack abilities can generate a massive amount of data with so many characteristics, such as a huge number of samples, many new attack types, and imbalanced data distribution Those problems are prevalent in the current cyber world, which undoubtedly reduces the performance of IDS. In order to help improve the detection accuracy and solve these problems, a method is proposed by us in this paper It combines the incremental extreme learning machine [9] (I-ELM) with adaptive principal component analysis (A-PCA) as our IDS’s detection algorithms.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
