An Improved Feature Selection Algorithm with Cyclic Penalty Boundary Interaction Based on MOPSO

Feature selection can minimize redundant features and improve classification performance. It is a multi-objective optimization problem for it has several conflicting objectives, and improving the multi-objective optimization algorithm is of great significance for feature selection. The Pareto mechanism cannot control the direction, but the decomposition mechanism can. However, its penalty factor is extremely to set. This paper proposes a novel feature selection algorithm based on multi- objective particle swarm optimization with cyclic penalty factor setting in PBI. To evaluate the performance of our proposed algorithm, six datasets are studied in our experiments. The experimental results demonstrate the superiority of the proposed algorithm over the comparison algorithms in obtaining the feature subsets with higher qualities.