Cluster Analysis of the Self-adaptive Ant Colony Algorithm with Crossover Operator

江波,李元香
DOI: https://doi.org/10.3321/j.issn:1000-565X.2004.z1.028
2004-01-01
Abstract:Data cluster is a key problem in the data mining, and the clustering can be regarded as an optimization problem. On the basis of Ant Colony Algorithm, a novel multi-colony clustering method is proposed in order to overcome the problems existing in the traditional clustering algorithm. In this method, the clustering principles followed by the human mind are adopted and the actions of ants when looking for foods are simulated. Simulation results show that the proposed method is of the advantages of high parallel, high efficiency and strong clustering and learning ability. The method can be used to obtain the optimal value of the whole field.
What problem does this paper attempt to address?