Parallel Clustering on a Commodity Supercomputer

Giuseppe Patanè and Marco Russo

Abstract

K-means based clustering algorithms have interesting performances in several application fields. The computational complexity of these techniques depends on the size of the data set and the codebook. The larger the data set and the codebook, the greater the computing time to reach the convergence. This paper illustrates the behaviour of two clustering algorithms we have realized and parallelized on a commodity supercomputer.

Keywords: Parallel Clustering, Unsupervised Learning, GLA, LBG, ELBG.

[MyIJCNN00] G.Patanè and M. Russo, Parallel Clustering on A Commodity Supercomputer, in IJCNN 2000, In Proc. of the IEEE-INNS-ENNS Int. Joint Conf. on Neural Networks, vol. 3, pp 575-580, 2000.

Full paper