An introduction to the use of probabilistic techniques in the design of expert systems, the book provides a background in probability and graph theory. It shows how to use the probability calculus in modelling reasoning under uncertainty, discusses subjective Bayesian analysis and the odds method and covers the theoretical foundations of causal networks, including Pearl's method of belief propagation and fusion in causal networks. Each chapter is split into two sections - the first section presents the theory and the second explains the technique's use and offers examples.
Sprache
Verlagsort
Verlagsgruppe
Zielgruppe
Für höhere Schule und Studium
Für Beruf und Forschung
Illustrationen
Maße
Gewicht
ISBN-13
978-0-471-61840-9 (9780471618409)
Copyright in bibliographic data is held by Nielsen Book Services Limited or its licensors: all rights reserved.
Schweitzer Klassifikation
Background: probabilistic considerations; graph theoretic considerations; rule based expert systems: subjective Bayesian method; the odds method; causal networks: the use of causal networks in expert systems; fusion, propagation and structuring in causal networks; "marrying" and "filling in" in causal networks; maximum entropy methods in causal networks; abductive inference - the set covering method; the probabilistic causal model; the odds method applied to set covering; higher order probabilities - the use of intervals instead of point values; the use of causal networks to express uncertainty in point values.
