Feature Selection with a Backtracking Search Optimization Algorithm

Feature selection carries significance in the outcome of any classification or regression task. Exercising evolutionary computation algorithms feature has led to construction efficient discrete optimization algorithms. In this paper, a modified backtracking search algorithm is employed perform wrapper-based selection, where two modifications standard are adopted. The first one concentrates on utilizing particle ranking operator regarding current population. second focuses removing case using single mutation process. Then, implementation above carried out terms general frameworks, which originally were developed for swarm optimization. framework based binary and set-based experimental analysis shows that variants equally well several datasets.