Towards a Machine Learning Oriented Expert System for Intrusion Detection Model

Abstract

Abstract In this paper, we describe the possibility of using machine learning and expert system for constructing a new intrusion detection model. The first main idea is to construct the classifier on the KDD intrusion data using a machine learning algorithm. For constructing the classifier J48 decision tree machine learning algorithm is used. The other key idea is to collect domain expert knowledge, and building an expert system to interpret and suggest possible countermeasures for the detected attack. We achieve an effective prediction and prevention from machine learning and expert system on a large collection of intrusion datasets. Keywords: Intrusion Detection, Machine Learning, Network attack, Expert SystemJ48 decision tree