Anomaly detection system using entropy based technique

An Intrusion detection system (IDS) is a module of software and/or hardware that monitors the activities occurring in a computer system or network system. The IDSs use various algorithms for detecting malicious activities. One of them is feature selection algorithm that depends on dimensionality reduction of the datasets. In this paper, we propose a novel feature selection algorithm based on information gain (entropy). We use the Knowledge Discovery and Data Mining cup dataset´99 for detecting the attacks and to classify them in four categories as well. Our algorithm provides better detection rate than the existing Fast Feature Reduction in Intrusion Detection Datasets (FFRIDD) and Multi-Level Dimensionality Reduction Methods (MLDRM).