Implementation of Machine Learning Algorithms on CICIDS-2017 Dataset for Intrusion Detection Using WEKA
Shailesh Singh Panwar1, Pritam Singh Negi2, Lokesh Singh Panwar3, Y. P. Raiwani4
1Shailesh Singh Panwar, Department of Computer Science and Engineering, H.N.B. Garhwal University Srinagar Garhwal, Uttarakhand, India,
2Pritam Singh Negi, Department of Computer Science and Engineering, H.N.B. Garhwal University Srinagar Garhwal, Uttarakhand, India.
3Lokesh Singh Panwar, Department of Electronics and Communication Engineering, H.N.B. Garhwal University Srinagar Garhwal, Uttarakhand, India.
4Y. P. Raiwani*, Department of Computer Science and Engineering, H.N.B. Garhwal University Srinagar Garhwal, Uttarakhand, India.
Manuscript received on 9 August 2019. | Revised Manuscript received on 18 August 2019. | Manuscript published on 30 September 2019. | PP: 2195-2207 | Volume-8 Issue-3 September 2019 | Retrieval Number: C4587098319/2019©BEIESP | DOI: 10.35940/ijrte.C4587.098319
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: For protecting and securing the network, with Intrusion Detection Systems through hidden intrusion has become a popular and important issue in the network security domain. Detection of attacks is the first step to secure any system. In this paper, the main focus is on seven different attacks, including Brute Force attack, Heartbleed/Denial-of-service (DoS), Web Attack, Infiltration, Botnet, Port Scan and Distributed Denial of Service (DDoS). We rely on features derived from CICIDS-2017 Dataset for these attacks. By using various subset based feature selection techniques performance of attack has been identified for many features. Using these techniques, it has been determined the appropriate group of attributes for finding every attack with related classification algorithms. Simulations of these techniques present that unwanted feature can be removed from attack detection techniques and find the most valuable set of attributes for a definite classification algorithm with discretization and without discretization, which improve the performance of IDS.
Keywords: IDS, CICIDS-2017, Classification Algorithms, Features Selection, WEKA.
Scope of the Article: Artificial Intelligence and Machine Learning