Feature selection for intrusion detection systems

TL;DR

This paper reviews existing feature selection techniques for intrusion detection, proposes a new method handling continuous and discrete data, and demonstrates its effectiveness with a machine learning system achieving 99.9% accuracy.

Contribution

It introduces a novel feature selection method tailored for intrusion detection that outperforms benchmark methods.

Findings

Proposed method performs well against benchmarks.

Achieves 99.9% accuracy in distinguishing DDoS from benign traffic.

Provides insights for designing automated intrusion detection systems.

Abstract

In this paper, we analyze existing feature selection methods to identify the key elements of network traffic data that allow intrusion detection. In addition, we propose a new feature selection method that addresses the challenge of considering continuous input features and discrete target values. We show that the proposed method performs well against the benchmark selection methods. We use our findings to develop a highly effective machine learning-based detection systems that achieves 99.9% accuracy in distinguishing between DDoS and benign signals. We believe that our results can be useful to experts who are interested in designing and building automated intrusion detection systems.