Application Research On Data Ming Using Ant Colony Algorithm

An Ant Colony Optimization algorithm is essential a system based on agents that simulate the natural behavior of ants, including cooperation and adaptation. The Ant Colony Algorithm is at the start stage in the data mining.It is collateral, easy realization and combine with other method that An Ant Colony Optimization algorithm is applied more and more widely. Due to it wide application,Classification is an important subject of data mining.This paper proposes an algorithm, which is based on ant colony algorithm, for mining classification rule from categorical database.The algorithm consists of four steps,comprising rule construction,rule pruning,pheromone uodating and classification with new rule.Moreover, this article also proposes an algorithm,which is based on ant colony algorithm,for clustering rule from database.The algorithm is tested and performed,It has higher predictive accuracy.