Advanced Methods of Approximate Reasoning
Abstract
This research was directed toward establishing basic conceptual foundations for approximate reasoning concepts and toward the development of frameworks that facilitate the development and comparison of applicable techniques. Approximate reasoning is the common name utilized to to describe automated techniques for the representation and manipulation of imprecise, uncertain, unreliable, and vague information. Our attention was focused primarily on the development of conceptual bases for possibilities or fuzzy logic. Using a conceptual framework, previously employed to explain the meaning of the Dempster-Shafer calculus of evidence (i.e., possible-world semantics), we developed a semantic model for possibilities logics that clearly shows them to be substantially different from their probabilistic counterparts both in meaning and in formal structure.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 30, 1990
- Accession Number
- ADA232140
Entities
People
- Enrique H. Ruspini
Organizations
- SRI International