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

Classification for high-dimensional data is a challenging task due to a great number of redundant and irrelevant features. Genetic Programming (GP) has its built-in feature selection characteristics and is suitable for processing high-dimensional data. However, if the number of features in GP individual is not restricted, the bloat phenomenon will occur, which affects the generalization of the training model. Moreover, GP is easy to fall into local search in the evolutionary process. In this paper, a multi-objective multi-population GP classifier construction method is proposed, which uses a multipopulation co-evolution strategy to divide the evolving population into the main and auxiliary populations with different evolutionary strategies and different evaluation criteria, and employs a multiobjective strategy to ensure classification performance and restrict the number of features at the same time. Experiments on seven high-dimensional datasets show that our proposed multi-population co-evolution strategy and multi-objective strategy are all effective to improve the classification performance of GP classifiers. Comparisons with three state-of-the-art GP classifier construction methods show that our proposed methods achieve better or comparable classification performance on most cases.