Building expert systems by a modified perceptron network with rule-transfer algorithms
- 1 January 1990
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 77-82 vol.2
- https://doi.org/10.1109/ijcnn.1990.137698
Abstract
A sequential perceptron-like inference network with novel rule-transfer algorithms is proposed. This network provides an efficient method for building knowledge-based or expert systems. The model is derived from probability and fuzzy set theory. The model has been utilized to organize and maintain the knowledge based. Some critical problems such as the exclusive-OR limitation and the conditional independence assumption have been carefully considered in the modificationsKeywords
This publication has 1 reference indexed in Scilit:
- Fuzzy setsInformation and Control, 1965