摘要： 层次聚类算法是一类重要的聚类分析方法。传统的层次聚类算法的时间和空间复杂度很大，这使得聚类分析在大型数据集上的应用受到限制。该文提出一种基于重叠区的3阶段改进算法，该算法将大大减少算法的时间复杂度和空间复杂度。

关键词: 聚类分析, 数据对象, 层次聚类算法, 重叠区

Abstract: A prominent clustering algorithm is hierarchical clustering. But the time complexity and space complexity of the traditional hierarchical clustering are high,and this limits its use for large data set. This paper proposes a 3-phase algorithm based on overlapping partitions to reduce the time and memory requirement.

中图分类号: 

• TP311.13