Advanced Concepts and Methods of Approximate Reasoning
Abstract
The major portion of the research was devoted to the development of unified framework for the description of approximation reasoning methods that facilitates the study of their fundamental characteristics. This objective was attained by consideration of structures, defined in spaces of possible worlds that measure either the relative size of certain subsets (for probabilistic methods) or the similarity between possible states (for probabilistic methods). Possible worlds are formalizations of the notion of possible state or behavior of a system. Using this concept, an approximate reasoning problem may be described as one where available evidence is insufficient to determine if the actual state of the world lied among those conceivable possibilities, conceivable where a statement about the system is true. Keywords: Artificial intelligence.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1989
- Accession Number
- ADA216474
Entities
People
- Enrique H. Ruspini
Organizations
- SRI International