Comparison towards of Integration of Machine Learning Methods for Intrusion Detection Systems

Abstract

It is important to incorporate modern approaches in that sequence to enhance the efficiency and quality of computer attacks identification. Recent years, machine learning methods are widely applied in Intrusion Detection Systems (IDS). We propose in this study compares two machine learning methods, namely Support Vector Machine (SVM) and Reinforcement Learning (RL). An analysis of existing techniques and their comparison regarding speed and precision, in addition to other factors may aid future researchers in understanding the recent advancements in IDS field as well as in creating innovations to satisfy needs and requirements in terms of computer security. The experimental results using the intrusion detection from NSL-KDD dataset show that the proposed integration is well suited for enhancing IDS performances.