Feature Selection with Intelligent Dynamic Swarm and Fuzzy Rough Set

A feature selection method based on intelligent dynamic swarm and fuzzy rough set is proposed in this paper, in which the fitness function is dependency measure. The proposed method can identify irrelevant and redundant features, and after dropping them produces the reduced feature set. Results found by this method have been compared with the results obtained using another established methods of particle swarm optimization for the same datasets. Results of the performed experiments shows that the proposed algorithm is able to select the best set of features and is faster than the particle swarm optimization method.