An improved feature selection method based on ant colony optimization (ACO) evaluated on face recognition system

• Published in: Applied mathematics and computation Vol. 205; no. 2; pp. 716 - 725
• Main Authors: Kanan, Hamidreza Rashidy, Faez, Karim
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Amsterdam Elsevier Inc 15.11.2008; Elsevier
• Subjects: Ant colony optimization (ACO); Calculus of variations and optimal control; Exact sciences and technology; Face recognition; Feature selection; Genetic algorithm (GA); Mathematical analysis; Mathematics; Numerical analysis; Numerical analysis. Scientific computation; Numerical methods in mathematical programming, optimization and calculus of variations; Sciences and techniques of general use; Genetic algorithm (GA); Feature selection; Face recognition; Ant colony optimization (ACO); Optimization method; Shortest path; Pattern recognition; Implementation; Experimental result; Numerical analysis; Algorithm performance; Genetic algorithm; Applied mathematics; Database; Mathematical programming
• ISSN: 0096-3003; 1873-5649
• DOI: 10.1016/j.amc.2008.05.115

Feature selection (FS) is a most important step which can affect the performance of a pattern recognition system. This paper proposes a novel feature selection method based on ant colony optimization (ACO). ACO algorithm is inspired of ant’s social behavior in their search for the shortest paths to food sources. Most common techniques for ACO-based feature selection use the priori information of features. However, in the proposed algorithm classifier performance and the length of the selected feature vector are adopted as heuristic information for ACO. So, we can select the optimal feature subset in terms of shortest feature length and the best performance of classifier. The experimental results on face recognition system using ORL database show that the proposed approach is easily implemented and without any priori information of features, its total performance is better than that of GA-based and other ACO-based feature selection methods.