An improved ant swarm algorithm for clustering analysis of data mining

Data Mining Technology from the huge amount of data found in the potential of useful information and knowledge, Ant colony algorithm in dealing with combinatorial optimization problem has also been made a better results, this makes the ant colony algorithm applied to data mining prospects. In order to overcome the defects of K-means algorithm such as the local optima and sensitivity to initialization and noise data, a new Ant-Cluster algorithm is put forward in this paper. The algorithm uses the capacity of global search in ant algorithm, and solves the problems of K-means. The experiment shows that the algorithm is correct, efficient and fast, and increases the convergence speed.