A hybrid particle swarm optimization for feature subset selection byintegrating a novel local search strategy

Abstract—tFeature selection has been widely used in data mining and machine learning tasks to make a modelwith a small number of features which improves the classifier’s accuracy. In this paper, a novel hybridfeature selection algorithm based on particle swarm optimization is proposed. The proposed methodcalled HPSO-LS uses a local search strategy which is embedded in the particle swarm optimization toselect the less correlated and salient feature subset. The goal of the local search technique is to guidethe search process of the particle swarm optimization to select distinct features by considering theircorrelation information. Moreover, the proposed method utilizes a subset size determination schemeto select a subset of features with reduced size. The performance of the proposed method has beenevaluated on 13 benchmark classification problems and compared with five state-of-the-art featureselection methods.< final year projects >