Ant Colony Optimization for Multi-Label Correlation-based Feature Selection Method

This paper proposes a new ant colony optimization for Multi-Label Correlation-Based Feature Selection (ACO-ML-CFS). This ACO performs a global search within the space of candidate features to identify an optimal subset. The selected feature subset is used by a Multi-Label k-Nearest Neighbors algorithm (ML-kNN), in this paper. The performance of ACO-ML-CFS is evaluated and compared with our previously works; Hill-Climbing algorithm for Multi-Label Correlation-Based Feature Selection (HC-ML-CFS) and Genetic algorithm for Multi-Label Correlation-Based Feature Selection (GA-ML-CFS) across four benchmark multi-label datasets.