An Incremental Model for Effective Intrusion Detection Using Fuzzy Rough Set and Fuzzy ARTMAP

摘要

Intrusion Detection Systems (IDSs) examine more features to detect the known and unknown attacks in computer network. Very few feature selection methods may be redundant or contribute something to the detection process, do not achieve greater detection accuracy and also not time consume. For reducing the time taken and achieving better classification accuracy through feature selection process, we introduced a new incremental model for effective intrusion detection. This proposed model is combining an incremental feature selection algorithm and classification algorithm. We propose, a new Incremental Feature Selection Algorithm called CRF and Fuzzy Rough Set based Feature Selection algorithm (IFSA) and the existing classification algorithm called Fuzzy ARTMAP Neural Classifier. The experimental results of the proposed model show that this model detects the intruders with less time and high detection rate when tested with KDD'99 cup data set.