A Novel Feature Selection Approach Based on Swarm Intelligence

The computational complexity of a texture classification algorithm is limited by the dimensionality of the feature space. A feature selection algorithm that can reduce the dimensionality of problem is often desirable, which has been studied by many authors because of its impact on the complexity of classifiers, Furthermore, feature selection in high dimension space is a NP hard problem. This paper presents a novel approach to solve feature subset selection based on improved ant colony optimization algorithm which hybrids heuristics information. The proposed approach has been implemented and tested on a real image texture classification problem. The results of proposed method are encouraging and outperform that of the presented ant colony optimization algorithm without heuristic information in this domain.