An Improved RF based Classification Method For Network Intrusion Detection System

  • K. NandhaKumar

Abstract

The growth of cyber security is a major concern that is due to increase in numbers of network digital devices and boundless communication pattern. New attack types are discovered by intruders in a day to day basis and these are identified properly for the prevention using the intrusion detection systems (IDS) for giving proper response to it. Three main elements are included in IDS that plays a crucial role in network security is feature conversion/selection, data collection, and decision engine. The systems efficiency is affected directly by the third component named decision engine and this issue can overcome by using some machine learning methods which is a promising research areas. In this paper, a new method named improved Random Forest method is used for big data to obtain high accuracy and low training time with its discriminative mechanism. Constructive research methods are adopted over this work. The test results show that the proposed strategy is giving an increasingly effective in classification process on test dataset. The parameters which used for the evaluation are Precision, Recall, and Accuracy the proposed method proves to be performing better.

Published
2019-12-21
How to Cite
NandhaKumar, K. (2019). An Improved RF based Classification Method For Network Intrusion Detection System. International Journal of Advanced Science and Technology, 28(17), 192 - 200. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/2244