Nearest neighbor clustering algorithm based on AIC criterion
Qin Xuan-yun · Systems engineering and electronics · 2005
The difficulty of cluster analysis lies in determining sort centers and sort numbers for a clustering model. The nearest neighbor clustering rule is given first and the sorting method based on this rule is set up. Then AIC criterion is proposed for different models. At last, a theoretical analysis is made for solving the contradiction between compactness of clustering and increased sort numbers. By means of an instance simulation, this method is proved to be correct and valuable in applications.