Overview feature selection using fish swarm algorithm

Abstract

Feature selection is a process of representing wanted features based on the requirement needed by selecting the best subset of a dataset without changing the originality of the dataset. The aim of feature selection is to obtain most optimal feature subset to represent the data and for that purpose feature selection offered a few methods. This paper gives an easy understanding of the feature selection concept and the available methods in feature selection. As nowadays metaheuristics is catching attention researchers in many fields and feature selection is one of them, this paper intentionally brief feature selection using metaheuristics that implement Fish Swarm Algorithm (FSA) in the feature selection process. FSA classified as one of the Swarm Intelligence (SI) techniques have several advantages mainly to solve optimization problems. A number of previous works are reviewed. Based on the reviewed and the outcome results that has been tested using high dimensional, real-valued benchmark data sets, FSA reflect good performance among others SI.