摘要
提出了利用凸包思想求取障碍距离的方法,并对该方法进行了理论与实践上的探讨。然后以所求得的障碍距离为依据,利用改进后的k-中心点法进行实验。实验表明,其结果不仅可以得到两点之间的障碍距离与障碍路径,而且可以实现顾及障碍物聚类的要求。最后对实验结果做出了总结与展望。
The clustering analysis is an important way to capture information in data mining. The method of calculating obstacle distance was brought forward based on the principle of convex hulls in this paper, and was discussed theoretically and practically. Then according to the obstacle distance, the experiment by κ-medoids algorithm was improved. The result of the experiment showed not only the obstacle distance and route between two points, but also filling the request of clustering considering obstacle. Finally, the sum-up and prospect was carried out according to the result of the experiment that was done by this method.
出处
《测绘科学技术学报》
北大核心
2008年第2期145-148,共4页
Journal of Geomatics Science and Technology
关键词
聚类分析
凸包
障碍距离
κ-中心点法
clustering analysis
convex hull
obstacle distance
κ-medoids algorithm
