Two strategies are presented for the parallelisation of the k-means clustering algorithm on MIMD distributed-memory computers (multicomputers). The implementations are decribed in detail The advantages of each outlined.
Keywords: Speech recognition, VQ Codebook, k-means clustering, parallel processing
Bibliographic reference. Elliott, J. A. / Forsyth, M. E. / McInnes, F. R. / Ramsey, N. W. (1993): "Parallelising k-means clustering on distributed memory MIMD computers", In EUROSPEECH'93, 1627-1630.