Feature Selection Using an Improved Multi-objective Immune Algorithm for Intrusion Detection

Feature selection is important for running data classification. By selecting a suitable subset of features, the training time of the classification model can be reduced while the classification accuracy can be improved. In this paper, we propose a feature selection method using an improved multi-objective immune algorithm for intrusion detection. We modified a traditional multi-objective immune algorithm using an elite selection strategy based on the reference vectors, which not only can solve the imbalanced classification problem, but also can have a faster convergence speed. Experimental results on the NSL-KDD dataset show the higher classification accuracy of the proposed algorithm.