Statistical Pattern Recognition
This paper presents the basic concepts of statistical pattern recognition as a formalization of the common-sense approach to comparing an unclassified input pattern to a standard pattern or “template.” The two most basic problems of statistical pattern recognition are how to measure goodness of fit and how to create templates. Statistical concepts provide answers to both questions. A simple Gaussian model is used to obtain a solution that is optimal under certain assumptions. Although limited to relatively easy problems, this model shows how statistical methods provide a firm mathematical foundation for more advanced techniques.
Please disable any pop-up blockers for proper viewing of this Whitepaper.