ON K-MEDOID CLUSTERING OF LARGE DATA SETS WITH THE AID OF A GENETIC ALGORITHM - BACKGROUND, FEASIBILITY AND COMPARISON

A novel approach to the problem of k-medoid clustering of large data s ets is presented, using a genetic algorithm. Genetic algorithms compri se a family of optimization methods based loosely upon principles of n atural evolution. They have proven to be especially suited to tackle c omplex, large-scale optimization problems efficiently, including a rap idly growing variety of problems of practical utility. Our pilot study lays emphasis on the feasibility of GCA - our genetic algorithm for k -medoid clustering of large datasets - and provides some background in formation to elucidate differences with traditional approaches. The ex perimental part of this study is done on the basis of artificial data sets and includes a comparison with CLARA - another approach to k-medo id clustering of large data sets. introduced recently. Results indicat e that GCA accomplishes a better sampling of the combinatorial search space.