A feature selection method with feature ranking using genetic programming

Feature selection is a data processing method which aims to select effective feature subsets from original features. based on evolutionary computation (EC) algorithms can often achieve better classification performance because of their global search ability. However, methods using EC cannot get rid invalid features effectively. A small number still exist till the termination algorithms. In this paper, genetic programming (GP) combined with ranking (FRFS) proposed. It assumed that more appear in GP individuals' terminal nodes, valuable these are. To further decrease selected features, FRFS multi-criteria fitness function named as MFRFS investigated. Experiments 15 datasets show obtain higher smaller compared without ranking. reduces while maintaining FRFS. Comparisons five benchmark techniques performance.