Dynamic clustering for interval data based on L2 distance
Dynamic clustering for interval data based on L2 distance
复制标题
DOI：
10.1007/s00180-006-0261-z
复制
发表时间：
2006-01-01
影响因子：
1.3
通讯作者：
Bock, Hans-Hermann
中科院分区：
文献类型：
--
作者：
de Carvalho, Francisco de A. T.;Brito, Paula;Bock, Hans-Hermann
本文介绍了一种区间数据描述对象的分区聚类方法。它遵循动态聚类方法并使用L-2距离。重点讨论了区间变量的标准化问题，提出并研究了区间变量的三种标准化技术。此外，还介绍了各种聚类解释工具，并通过模拟和实际案例数据进行了说明。
This paper introduces a partitioning clustering method for objects described by interval data. It follows the dynamic clustering approach and uses an L-2 distance. Particular emphasis is put on the standardization problem where we propose and investigate three standardization techniques for interval-type variables. Moreover, various tools for cluster interpretation are presented and illustrated by simulated and real-case data.