다차원 개체를 위한 차이 등급 clustering

Abstract

The clustering problem for multi-dimensional entities is investigated. A heuristic method, which is named as Difference Order Clustering (DOC), is developed for the grouping of multi-dimensional entities. DOC method has an advantage of identifying the bottle-neck entities. Comparisions among the proposed DOC method, modified rank order clustering (MODROC) method, and lexicographical rank order clustering using minimum spanning tree (lexico-MMSTROC) are illustrated by a part type selection problem.