Reduct ECOC Framework for Network Intrusion Detection System
Uma Shankar Rao Erothi1, Sireesha Rodda2
1Uma Shankar Rao Erothi*, Department of CSE, RAGHU Institute of Technology, Visakhapatnam, India.
2Sireesha Rodda, Department of CSE, GITAM University, Visakhapatnam, India.
Manuscript received on February 01, 2020. | Revised Manuscript received on February 05, 2020. | Manuscript published on February 30, 2020. | PP: 258-266 | Volume-9 Issue-3, February, 2020. | Retrieval Number: B4238129219/2020©BEIESP | DOI: 10.35940/ijeat.B4238.029320
Open Access | Ethics and Policies | Cite | Mendeley
© 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: : Now a day’s network security is major concern for e-government and e-commerce applications. A wide range of malicious activities are increasing with the usage of internet and network technologies. Identifying novel threats and finding modern solutions for network to prevent from these threats are important. Designing an effective intrusion detection system is significant to continuously look out the network activities to efficiently thwart malicious attacks or to identify the intruders. To tackle multi class imbalance classification problem in networks, a reduct based ECOC ensemble framework for NIDS is proposed to efficiently identify attacks in a multi class scenario. The Reduct-ECOC classifier is validated on highly imbalanced benchmark NSL-KDD intrusion datasets as well as other UCI-ML datasets. The experimental results on eight highly imbalanced datasets show that Reduct-ECOC classifier performs better than many other state-of-art multi-class classification ECOC learning methods.
Keywords: Network Intrusion Detection System, Multi Class Imbalance, Rough Set Theory, ECOC Classifier.