APPLICATION OF UNCERTAIN VARIABLES IN A CLASS OF KNOWLEDGE-BASED PATTERN RECOGNITION SYSTEMS

摘要

A concept of uncertain variables based on uncertain logics has been developed for a class of uncertain systems containing unknown parameters in their mathematical descriptions. The purpose of this paper is to show how this approach may be applied in a class of pattern recognition systems described by relational knowledge representations. A short presentation of the uncertain variables is given. A pattern recognition problem with uncertain parameters in the knowledge representation is considered and a procedure of finding the optimal classification is described. A simple example illustrates the presented approach.