摘要
本文在分析现有的基于网格和密度的聚类方法的基础上,借鉴密度函数的思想,提出了一种新的网格单元密度的计算方法。和其他基于网格的方法相比,此方法对于网格单元密度的计算具有更高的精度。在此基础上,将该计算方法与微粒群算法相结合,实现了一种基于网格和密度的微粒群混合聚类算法。最后,通过实验验证了算法的有效性。
The purpose of the paper is to present a new method for computing the density of the grid. Furthermore, a grid density-based clustering algorithm with the hybridization of the Particle Swarm Optimization(PSO) is also presented. The method for computing the grid-density shown in the paper can be more precise than most of the others. Finally, the results of the experiments indicate the efficiency of the algorithm.
出处
《计算机科学》
CSCD
北大核心
2006年第11期164-165,187,共3页
Computer Science
基金
国家自然科学基金资助项目:70272050