Intrusion Detection Using Fuzzy Stochastic Local Search Classifier

This paper proposes a stochastic local search classifier combined with the fuzzy logic concepts for intrusion detection. The proposed classifier works on knowledge base modeled as a fuzzy rule "if-then" and improved by using a stochastic local search. The method is tested on the Benchmark KDD\´99 intrusion dataset and compared with other existing techniques for intrusion detection. The results are encouraging and demonstrate the benefit of the proposed approach.