MUTUAL INFORMATION BASED FEATURE SELECTION TECHNIQUES FOR INTRUSION DETECTION

Abstract: A huge amount of high dimensional audit data is the major problem for accurate & quick detection of the intrusions. The audit data may contains some irrelevant & redundant features. Processing of these features by an IDS may increase the computational overhead, decrease the overall accuracy, and delay the process of intrusion detection. Therefore, for accurate & quick intrusion detection, the audit data may be reduced by selecting the most relevant and non-redundant features.In this paper, we explored various feature selection techniques especially mutual information (MI) based filter feature selection techniques. An updated review of the important techniques in literature is presented. The review will help the better understanding of different directions in which research has been done in the field of feature selection. The findings of this paper provide useful insights into literature and are beneficial for those who are interested in applications of MI based feature selection techniques to IDS and related fields. The review also provides the future directions of the research in this area.

Keywords: Feature reduction Data Breaches Feature selection Intrusions Intrusion detection Network Security Security Threats1