From the Publisher:
Probabilistic Reasoning in Intelligent Systems is a complete andaccessible account of the theoretical foundations and computational methods that underlie plausible reasoning under uncertainty. The author provides a coherent explication of probability as a language for reasoning with partial belief and offers a unifying perspective on other AI approaches to uncertainty, such as the Dempster-Shafer formalism, truth maintenance systems, and nonmonotonic logic. The author distinguishes syntactic and semantic approaches to uncertaintyand offers techniques, based on belief networks, that provide a mechanism for making semantics-based systems operational. Specifically, network-propagation techniques serve as a mechanism for combining the theoretical coherence of probability theory with modern demands of reasoning-systems technology: modular declarative inputs, conceptually meaningful inferences, and parallel distributed computation. Application areas include diagnosis, forecasting, image interpretation, multi-sensor fusion, decision support systems, plan recognition, planning, speech recognitionin short, almost every task requiring that conclusions be drawn from uncertain clues and incomplete information.
Probabilistic Reasoning in Intelligent Systems will be of special interest to scholars and researchers in AI, decision theory, statistics, logic, philosophy, cognitive psychology, and the management sciences. Professionals in the areas of knowledge-based systems, operations research, engineering, and statistics will find theoretical and computational tools of immediate practical use. The book can also be used as an excellent text for graduate-level courses in AI, operations research, or applied probability.

Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference

Possible solutions to the problem of combining classifiers can be divided into three categories according to the levels of information available from the various classifiers. Four approaches based on different methodologies are proposed for solving this problem. One is suitable for combining individual classifiers such as Bayesian, k-nearest-neighbor, and various distance classifiers. The other three could be used for combining any kind of individual classifiers. On applying these methods to combine several classifiers for recognizing totally unconstrained handwritten numerals, the experimental results show that the performance of individual classifiers can be improved significantly. For example, on the US zipcode database, 98.9% recognition with 0.90% substitution and 0.2% rejection can be obtained, as well as high reliability with 95% recognition, 0% substitution, and 5% rejection. >

Methods of combining multiple classifiers and their applications to handwriting recognition

The use of belief functions to represent and to manipulate uncertainty in expert systems has been advocated by some practitioners and researchers. Others have provided examples of counter-intuitive results produced by Dempster's rule for combining belief functions and have proposed several alternatives to this rule. This paper presents another problem, the failure to balance multiple evidence, then illustrates the proposed solutions and describes their limitations. Of the proposed methods, averaging best solves the normalization problems, but it does not offer convergence toward certainty, nor a probabilistic basis. To achieve convergence, this research suggests incorporating average belief into the combining rule.

Combining belief functions when evidence conflicts

The theory of evidence proposed by G Shafer is gaining more and more acceptance in the field of artificial intelligence, for the purpose of managing uncertainty in knowledge bases One of the crucial problems is combining uncertain pieces of evidence stemming from several sources, whether rules or physical sensors This paper examines the framework of belief functions in terms of expressive power for knowledge representation It is recalled that probability theory and Zadeh's theory of possibility are mathematically encompassed by the theory of evidence, as far as the evaluation of belief is concerned Empirical and axiomatic foundations of belief functions and possibility measures are investigated Then the general problem of combining uncertain evidence is addressed, with focus on Dempster rule of combination It is pointed out that this rule is not very well adapted to the pooling of conflicting information Alternative rules are proposed to cope with this problem and deal with specific cases such as nonreliable sources, nonexhaustive sources, inconsistent sources, and dependent sources It is also indicated that combination rules issued from fuzzy set and possibility theory look more flexible than Dempster rule because many variants exist, and their numerical stability seems to be better



La theorie des fonctions de credibilite proposee par G Shafer est de plus en plus reconnue, dans le domaine de l'intelligence artificielle, pour le traitement de l'incertitude dans les bases de connaissances Un des problemes cruciaux est la combinaison d'informations issues de plusieurs sources (regles ou capteurs) Cet article etudie le cadre propose par les fonctions de credibilite quant a son pouvoir expressif pour la representation de connaissances On rappelle que d'un point de vue mathematique la theorie des probabilites et la theorie des possibilites sont des cas particuliers de la theorie des fonctions de credibilite pour ce qui est de l'evaluation de l'incertitude Les fondements empiriques et axiomatiques des fonctions de credibilite et de possibilite sont discutes Puis le probleme general de la combinaison est aborde en mettant l'accent sur la regle de Dempster On met en evidence le fait que cette regle est peu adaptee a la combinaison d'informations conflictuelles D'autres regles sont proposees pour traiter ce probleme dans diverses situations telles que sources inegalement fiables, incompletes, incoherentes, dependantes On indique aussi que les regles de combinaison proposees par la theorie des ensembles flous et des possibilites semblent plus flexibles que la regle de Dempster, car de nombreuses variantes existent; de plus leur stabilite numerique semble meilleure

G. Shafer

Papers

Implementing Dempster's rule for hierarchial evidence