期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：7
出版社：Journal of Theoretical and Applied
摘要：Affinity Propagation is one of clustering technique that use iterative message passing and consider all data points as potential exemplars. It is complimented because providing a good result of clustering with low error rate. But it has several drawback, such as quadratic computational cost and vague values of preference. There are many research trying to solve the drawback to improve the speed and quality of Affinity Propagation. But, there are not any test to find the best Affinity Propagation expansion algorithm in speed. This has led researchers to try to compare the performance of several Affinity Propagation expansion algorithms. The tested algorithms are Adaptive Affinity Propagation, Partition Affinity Propagation, Landmark Affinity Propagation, and K-AP. There are two comparison made in this paper: theoretical analysis and running test. From both comparison, it can be found that Landmark Affinity Propagation has the most efficient computational cost and the fastest running time, although its clustering result is very different in number of clusters than Affinity Propagation
