AN EFFICIENT ANOMALY INTRUSION DETECTION METHOD WITH EVOLUTIONARY KERNEL NEURAL NETWORK RANDOM WEIGHTS | [Journal of Theoretical and Applied Information Technology • 2020]

Author(s):

1. SAMIRA SARVARI: Department of Computer Science, Faculty of Computer Science and Information Technology Universiti Putra Malaysia, UPM Serdang Selangor, MALAYSIA

2. NOR FAZLIDA MOHD SANI: Department of Computer Science, Faculty of Computer Science and Information Technology Universiti Putra Malaysia, UPM Serdang Selangor, MALAYSIA

3. ZURINA MOHD HANAPI: Department of Communication Technology and Network, Faculty of Computer Science and Information Technology, Universiti Putra Malaysia,43400 UPM Serdang Selangor, MALAYSIA

4. MOHD TAUFIK ABDULLAH: Department of Computer Science, Faculty of Computer Science and Information Technology Universiti Putra Malaysia, UPM Serdang Selangor, MALAYSIA

Abstract:

Internet security requirements are increasing due to the growth of internet usage. One of the most efficient approaches used to secure the usage of the internet from internal and external intruders is Intrusion Detection System (IDS). Considering that using a combination of ANN and EA can produce an advanced technique to develop an efficient anomaly detection approach for IDS, several types of research have used ENN algorithms to detect the attacks. To enhance the efficiency of anomaly-based detection in terms of accuracy of classification, in this paper, the evolutionary kernel neural network random weight is proposed. This model is applied to the NSLKDD dataset, an improvement of the KDD Cup'99. The proposed method achieved 99.24% accuracy which shows that the novel algorithm suggested is more superior to existing ones as it provides the optimal overall efficiency.

Page(s): 963-976

DOI: DOI not available

Published: Journal: Journal of Theoretical and Applied Information Technology, Volume: 98, Issue: 7, Year: 2020

Keywords:

Intrusion detection systems IDSs , NSLKDD Dataset , Multilayer perceptron MLP , Multiverse optimizer MVO

References:

References are not available for this document.

Citations

Citations are not available for this document.

Citations

Downloads

Views