A novel feature selection technique based on Roach Infestation Optimization for Internet Traffic Classification

Abstract

In this paper, we propose a novel feature selection technique based on the Roach Infestation Optimization (RIO) meta heuristic for Internet traffic classification. The RIO based feature selection technique is a pre-processing step before launching the classification task where the aim is to identify the set of significant features to be used in the classification task. The proposed technique is combined with both random forest and Bayes network classifiers and evaluated on the well known NIMS dataset. The numerical results show the effectiveness of the proposed technique for Internet traffic classification.