Empirical Comparison of Similarities for Agglomerative Hierarchical Clustering

This paper proposes a method for empirical comparison of distances for agglomerative hierarchical clustering based on rough set-based approximation. When a set of target is given, a level of clustering tree where one branch includes all the targets can be traced with the number of elements included. The pair (#clusters of a level, #elements of a cluster) can be viewed as indices-pair for a given clustering tree.

I. Introduction

Introduction of a hospital information system enables us to store all the clinical information and to apply machine learning and data mining methods to those data [1]–[3]. For example, ordinary statistical methods are applied for hospital management [4], [5], temporal data mining are applied for capturing behavior of medical staff and for analyzing disease progression [6]–[8].