The paper is concerned with static plants described by a knowledge representation in the form of relations or logical formulas with unknown parameters. The learning algorithms consisting of step‐by‐step knowledge validation and updating have been presented for different versions. The theorems concerning the convergence of the learning processes have been given for the open‐loop and closed‐loop learning decision‐making system. The extension for the case with uncertain observations has been suggested. Simple examples illustrate the presented approach.
CitationDownload as .RIS
MCB UP Ltd
Copyright © 2000, MCB UP Limited