ABSTRACT
CONTENTS 20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 20.2 Dissimilarity Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340 20.3 DCV Metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343 20.4 Synthetic Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 20.5 Exploiting the DCV Metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348 20.6 Conclusions and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
A measure of the closeness between two objects in high dimensions is essential to most clustering algorithms. In a metric space, the closeness between two objects is modelled by a distance function that satisfies the triangle inequality. Most earlier work on metrics focussed on continuous fields. However, many of the fields in data mining consist of categorical data which describe the symbolic values of objects. In this chapter, we introduce a new metric for categorical data that is derived from the Mahalanobis metric.
