Introduction and Literature Review

We live in a world where data are increasingly becoming available in larger quanti- ties than ever

before and increasingly form the basis of decisions in daily activities. To cope with the

ever"increasing volume of data, it is imperative to organize data effectively and efficiently.

The very first step of any data organization activity is to classify data. The present

thesis emphasizes three aspects of the classification of data. First, it takes a data mining

approach as opposed to an empirical or inferen- trial approach. Second, it restricts the

discussion only to sequence data, including multidimensional and time-series data. Third and

final, it focuses mainly on un- supervised classification, that is clustering.