Matching interval-valued-argument propositions in rule-based systems
CSC '90 Proceedings of the 1990 ACM annual conference on Cooperation
| Hi-index | 0.00 |
This paper describes an architecture for incorporating adaptive learning in rule-based diagnostic expert systems. An adaptive experience-based diagnostic expert system prototype is described which develops and maintains its knowledge base by using adaptive learning techniques. By coupling new experience with the current state of its knowledge, the system incrementally refines the rules in its knowledge base. Although the system described performs as an assistant in troubleshooting electronic component failures in aircraft, the inductive learning mechanisms used are applicable to other rule-based diagnostic systems which are located in environments capable of providing feedback to the system.