Feature Reduction using Lasso Hybrid Algorithm in Wireless Intrusion Detection System
D. Sudaroli Vijayakumar1, Sannasi Ganapathy2
1D.Sudaroli Vijayakumar, Computer Science, PES University, Bangalore, India.
2Sannasi Ganapathy, Computer Science, VIT University, Chennai, India.
Manuscript received on 20 August 2019. | Revised Manuscript received on 08 September 2019. | Manuscript published on 30 September 2019. | PP: 1476-1483 | Volume-8 Issue-11, September 2019. | Retrieval Number: J98100881019/2019©BEIESP | DOI: 10.35940/ijitee.J9810.0981119
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: To maintain the integrity and protection of networks, intrusion detection systems play a vital role. Growth of wireless networks turned the globe to perform all pecuniary tasks online resulting a lot of security breaches in the network. One of the common breaches happening in network is the intruders who eventually tries to bypass the adopted security framework. Every day new intrusions arises and new solutions as well, however the research in making the intrusion detection system intelligent holds energetic. Today most of the systems are becoming intelligent by adopting machine learning and artificial intelligence algorithms. Success of building an efficient machine learning model to make intelligent intrusion detection system is relied on the effective features considered for classification and prediction. Thus, feature reduction is an integral part for discarding irrelevant and redundant features to produce a computationally decisive system that can identify defects with high accuracy. This implementation is an attempt to identify the smaller feature set possible for the well adopted wireless intrusion detection dataset AWID. Here, we proposed a LASSO based implementation to produce a smaller decisive set of features. Incorporation of Lasso on feature reduction not only provides a smaller set of features, but also allow to adopt prediction algorithms inside Lasso resulting lesser number of false alarms as well.
Keywords: WLAN 802.11, Intrusion Detection System (IDS), Machine Learning, Enhanced Feature Selection (EFS), Random Forest, Lasso, Wireless IDS (WIDS).
Scope of the Article: Machine Learning