Stochastic Model for the Browning-Bledsoe Pattern Recognition Scheme

A stochastic model is presented which gives the probabilities of successful recognition of the Browning-Bledsoe recognition scheme as a function of scheme parameters and pattern variability parameters. Also, procedures are given for estimating the variability parameters from data so that the model can be used to predict readability. The adequacy of the model is checked by comparing estimated readability with observed readability for two sets of data, one with high variability and one with low variability. The Browning-Bledsoe recognition scheme is also treated as a coding and decoding problem in which case the concepts of information theory are useful. Finally, brief mention is made of the connection between pattern recognition problems and classification problems in general, and the Browning-Bledsoe recognition scheme is compared and contrasted with other recognition schemes which make use of measurements on patterns.