基于距离不等式的K-medoids聚类算法  

K-Medoids Clustering Algorithm Based on Distance Inequality

作　　者：余冬华[1] 郭茂祖[1,2] 刘扬 任世军[1] 刘晓燕[1] 刘国军[1] 

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,黑龙江哈尔滨150001 [2]北京建筑大学电气与信息工程学院,北京100044

出　　处：《软件学报》2017年第12期3115-3128,共14页Journal of Software

基　　金：国家自然科学基金(61571164;61571163;61671188;61671189;QC2014C071)~~

摘　　要：研究加速K-medoids聚类算法,首先以PAM(partitioning around medoids)、TPAM(triangular inequality elimination criteria PAM)算法为基础给出两个加速引理,并基于中心点之间距离不等式提出两个新加速定理.同时,以O(n+K^2)额外内存空间开销辅助引理、定理的结合而提出加速SPAM(speed up PAM)聚类算法,使得K-medoids聚类算法复杂度由O(K(n-K)~2)降低至O((n-K)~2).在实际及人工模拟数据集上的实验结果表明:相对于PAM,TPAM,FKMEDOIDS(fast K-medoids)等参考算法均有改进,运行时间比PAM至少提升0.828倍.This paper presents a research on speeding up K-medoids clustering algorithm. Firstly, two acceleration lemmas are given based on partitioning around medoids （PAM） and triangular inequality elimination criteria PAM （TPAM） algorithms. Then two new acceleration theorems are proposed based on distance inequality between center points. Combining the lemmas with the theorems with the aid of additional memory space O（n＋K^2）, the speed up partitioning around medoids （SPAM） algorithm is constructed, reducing the time complexity from O（K（n-K）2） to O（（n-K）^2）. Experimental results on both real-world and artificial datasets show that the SPAM algorithm outperforms PAM, TPAM and FKEMDOIDS approaches by at least 0.828 times over PAM in terms of running time.

关 键 词：数据挖掘 聚类算法 K-medoids 距离不等式 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]