Dissimilarity Computation for Objects of Different Variable Types

Abstract

Clustering is the process of grouping a set of physical or abstract objects into classes of similar objects is called clustering. A cluster is a collection of data objects that are similar to one another within the same cluster and are dissimilar to the object in other cluster. Measuring the dissimilarity between data objects is one of the primary tasks for distance-based techniques in data mining and machine learning, e.g., distance-based clustering and distance-based classification. The quality of clustering can be accessed based on dissimilarity measures of objects which can be computed for various types of data. In this paper, we propose general framework for measuring a dissimilarity betweens various data analysis is proposed. The key idea is to consider the dissimilarity between two values of an attribute as a combination of dissimilarities between the conditional probability distributions of other attributes given these two values. In this system, the similarity is guessed by computing the dissimilarity measure between two objects. This can get the most similar values and the least similar values before clustering analysis.