计算机应用研究2018，Vol.35Issue(2)：372-374,399,4.DOI:10.3969/j.issn.1001-3695.2018.02.012
基于初始偏向度的AP算法聚类性能优化研究
Research on optimization of AP algorithm performance based on initial preference
赵延龙 1滑楠1
作者信息
- 1. 空军工程大学信息与导航学院,西安710077
- 折叠
摘要
Abstract
This paper found the problem in the process of simulating the AP algorithm that the initial preference of class representative points had strong connection with clustering performance.In order to research the quantitative relationship deeply,this paper built the initial preference multiple simple target optimization model with constraint condition that limited the time-consumption to specific range so that it got the optimal value of initial preference which could improve the accuracy of clustering algorithm and reduce the time complexity effectively.Experimental results show that for the three classical standard datasets 4k2-far、wine and iris,the improved AP algorithm can reduce the time-consumption and at the same time improve the clustering accuracy compared to the existing algorithm,thus improve the clustering performance of AP algorithm.关键词
AP算法/初始偏向度/多重单目标优化/聚类性能Key words
AP algorithm/initial preference/multiple simple target optimization/clustering performance分类
信息技术与安全科学引用本文复制引用
赵延龙,滑楠..基于初始偏向度的AP算法聚类性能优化研究[J].计算机应用研究,2018,35(2):372-374,399,4.
