A Multi-Objective Evolutionary Algorithm for Multimedia Traffic Classification

This paper proposes a feature selection algorithm based on information gain rate and multi-objective evolutionary computing. The information gain rate is used to rank features and remove redundant ones for rapid dimensionality reduction. The top ranked features are used as the initial population by the adaptive evolutionary algorithm to search the optimal feature subset. The experimental results show that this method has higher classification accuracy as well as lower time complexity than the existing methods.