关键词: 测试用例集约简，聚类分析，K中心点算法，错误检测率

Abstract: According to the given testing objectives, test suite reduction aims to satisfy all testing requirements with the minimum number of test cases. This paper introduced the idea of K-medoids algorithm of cluster analysis to finish the test suite reduction. First, treat every test case as a node and find its similarity. And then, according to testing require-ments,choose test cases from test case suite. The results of the emulate programs prove that the method is effective and feasible.

Key words: Test suite reduction,Cluster analysis,K-medoids,Fault detection effectiveness