According to the requirement of the topography selection, the present spatial cluster technology and methods were studied deeply, a structural element was established. A new algorithm of 3-D spatial clustering based on mathematical morphology was presented. The process of spatial data mining was given, and an artillery position selection system was realized. The algorithm realized cluster of spatial objects through closing. It could not only complete 3-D spatial cluster at a time, but also handle cluster in-convex and complicated objects rapidly. On the basis of mathematical morphology, the algorithm was easy to implement its high performance parallel algorithm. Experiment results show that the algorithm can supply an effective solving scheme for the computer realization of the artillery position selection, and discover the potential relationships and rules in vast spatial database when it is applied to the development of the artillery position selection system, so it can help decision makers to make sensible decisions.
�ź���,��˼��. �����ھ�����װ�����Ͼ���֧��ϵͳ�е�Ӧ��[J]. �����Ӧ��,2007,28(1):79-82 Zhang Haichuan, Wang Sichang. Application of data mining technique in decision support system of equipment support[J]. Microcomputer Applications, 2007,28(1):79-82(in Chinese)
����ӱ,�ֽ�. ���ں˷����ľ����㷨����Ӧ��[J]. �������պ����ѧѧ��,2006,32(6):747-750 Ji Qiuying, Lin Jian. Clustering algorithm based on kernel methods and its application[J]. Journal of Beijing University of Aeronautics and Astronautics,2006,32(6):747-750(in Chinese)
Xu Rui,Wunsch II Donald. Survey of clustering algorithms[J].IEEE Transactions on Neural Networks.2005,16(3):645-678
Ng Raymond T, Han Jiawei. CLARANS:A method for clustering objects for spatial data mining[J].IEEE Transactions on Knowledge and Data Engineering.2002,14(5):1003-1016