L. A. Becker
Abstract
Instead of acquiring diagnostic rules about a manufacturing process directly from domain experts, one can acquire a deeper, model-based representation, and compile the diagnostic rules directly from it. A manufacturing process representation (MPR) is a deep (Harmon, 1985), functional representation (Davis, 1984; Chandrasekaran, 1985) which embodies a model of the given process. This representation includes knowledge about the ordering of process steps, as well as what they are designed to accomplish and what would make them fail or what errors they may trap. As indicated below, an MPR can be acquired either by a knowledge engineer (KE), who would interview an expert and make use of any available process design documents, or by an Intelligent Interrogator (II) program.
- Becker, L. A., and A. Kinigadner (1987). "Distributed diagnosis of execution errors in a manufacturing system," in Knowledge Based Systems for Engineering: Classification, Education, and Control. D. Sriram and R. A. Adey (Eds). Boston, MA: Computational Mechanics Publications, pp. 83-93.Google Scholar
- Becker, L. A., R. Bartlett, and M. Roy (1988). "Compiling diagnostic rules from a manufacturing process representation," in Artificial Intelligence in Engineering: Diagnosis and Learning, in J. Gero (Ed.), Southhampton: Computational Mechanics Publications, pp. 285-305.Google Scholar
- Chandrasekaran, B., and R. Milne (1985). Special Section on Reasoning About Structure, Behavior, and Function, SIGART Newsletter, No. 93, pp. 5-54. Google ScholarDigital Library
- Davis, R. (1984). "Diagnostic reasoning based on structure and behavior," Artificial Intelligence, Vol. 24, pp. 347-410. Google ScholarDigital Library
- Harmon, P., and D. King (1985). Expert Systems: Artificial Intelligence in Business, New York, NY: John Wiley. Google ScholarDigital Library
Index Terms
- Using simulation to compile diagnostic rules from a manufacturing process representation
Recommendations
Mining multi-level diagnostic process rules from clinical databases using rough sets and medical diagnostic model
IFSA'03: Proceedings of the 10th international fuzzy systems association World Congress conference on Fuzzy sets and systemsOne of the most important problems on rule induction methods is that they cannot extract rules, which plausibly represent experts' decision processes. On one hand, rule induction methods induce probabilistic rules, the description length of which is too ...
Mining diagnostic rules from clinical databases using rough sets and medical diagnostic model
Special issue: Medical expert systemsSince rule induction methods generate rules whose lengths are the shortest for discrimination between given classes, they tend to generate rules too short for medical experts. Thus, these rules are difficult for the experts to interpret from the ...
Extracting easily interpreted diagnostic rules
Diagnostic rules are formulated as fuzzy focal elements using the DempsterShafer and fuzzy set theories.A data-driven rule extraction algorithm is proposed that results in a set of rules which are easy to interpret by a human user (an expert).An ...
Comments