One Inference Engine for Expert Systems with Incomplete Information.
Abstract:
This paper concerns consideration on methods of inference engine constructing in diagnostic expert systems. The problem of correct processing of incomplete and unreliable information was under special consideration. For example of solution of this problem we describe the inference engine of expert system VibroExpert, that is built on the base of five-layer neural network. This network is compiled basing on rules in the beginning of system working.
