A conditional opposition-based particle swarm optimisation for feature selection

Because of the existence irrelevant, redundant, and noisy attributes in large datasets, accuracy a classification model has degraded. Hence, feature selection is necessary pre-processing stage to select important features that may considerably increase efficiency underlying algorithms. As popular metaheuristic algorithm, particle swarm optimisation successfully applied various approaches. Nevertheless, tends suffer from immature convergence low rate. Besides, imbalance between exploration exploitation another key issue can significantly affect performance optimisation. In this paper, conditional opposition-based proposed used develop wrapper selection. Two schemes, namely learning strategy are introduced enhance Twenty-four benchmark datasets validate approach. Furthermore, nine metaheuristics chosen for verification. The findings show supremacy approach not only obtaining high prediction but also small sizes.