Brian J. d'Auriol, Ph.D.

Publication Details
Return to d'Auriol's Publication Page


Brian J. d'Auriol, Virendra C. Bhavsar and Lev Goldfarb, ``Multi-Transputer Implementations of the Metric Approach to Pattern Recognition Using Weighted Levenshtein Distance'', Applications of Transputer 3; Volume II, T.S. Durrani, et al. (Eds.), IOS Press, Amsterdam, pp. 388-393, August 1991.

A software realization of the systolic array algorithm for the computation of Weighted Levenshtein Distance (WLD) in Occam is discussed and its various implementations on multi-transputer networks are proposed. The WLD computation forms a major block of the Metric Approach to Pattern Recognition. This approach is a model of machine learning which has a wide range of applications including speech and character recognition.

Last Updated: July 28, 2007