Ant Feature Selection Applied to Marble Classification

One of the most important techniques in data preprocessing for data mining is feature selection. Real-world data analysis, data mining, classification and modeling problems usually involve a large number of candidate inputs or features. This paper proposes an ant colony optimization (ACO) algorithm for the feature selection problem. The goal is to find the set of features that reveals the best classification accuracy using neural or fuzzy classifiers. The algorithm is tested in the marble classification problem. Several features have been extracted measuring marble visual properties, such as color, texture, shape and dimensions of their components. The obtained results show the effectiveness of the proposed method.