Ant Colony Optimization (ACO) is an exciting model, which can solve a variety of optimization problems, based on observing the cooperative behavior of social insects such as ants. Recently, it has been adapted to the generation of an ordered set of IF-THEN rules, producing categorical classification of unknown input data. The purpose of this paper is to confirm the usefulness of the new application of the Ant Colonization Algorithm to generation of classification rules. This new classification model presents certain advantages similar to those attributed to statistical methods and neural networks, i.e. resistance to noise and attribute interaction, as well as the ability to produce simple rules, useful in the real world. We have implemented the in Visual Basic. We have analyzed six public data sets, mostly in the medical domain.
展开▼
