Pattern recognition: Historical perspective and future directions

Pattern recognition is one of the most important functionalities for intell igent behavior and is displayed by both biological and artificial systems. Pattern recognition systems have four major components: data acquisition an d collection, feature extraction and representation, similarity detection a nd pattern classifier design, and performance evaluation. In addition, patt ern recognition systems are successful to the extent that they can continuo usly adapt and learn from examples; the underlying framework for building s uch systems is predictive learning, The pattern recognition problem is a sp ecial case of the more general problem of statistical regression; it seeks an approximating function that minimizes the probability of misclassificati on. In this framework, data representation requires the specification of a basis set of approximating functions. Classification requires an inductive principle to design and model the classifier and an optimization or learnin g procedure for classifier parameter estimation. Pattern recognition also i nvolves categorization: making sense of patterns not previously seen. The s ections of this paper deal with the categorization and functional approxima tion problems; the four components of a pattern recognition system; and tre nds in predictive learning, feature selection using "natural" bases, and th e use of mixtures of experts in classification, (C) 2000 John Wiley & Sons, Inc.