An algorithm for finding most likely explanations in valuation based systems

• Autorzy: Wierzchoń S. , Kłopotek M.
• Języki publikacji: EN

Abstrakty

EN

A method for finding a number of best explanations in so-called valuation based system is presented. Roughly speaking, the method allows to sort (decreasingly or increasingly) a function of many variables without explicit computation of values of this function. The only condition is that the function be decomposable, i.e. can be expressed as a combination of a number of low-dimensional functions called components. Two cases are considered: the combination operator has an inverse and a more elaborated case when the combination operator has no inverse.

Słowa kluczowe

EN

Bayesian networks; most probable configurations; probabilistic reasoning; valuation based systems

PL

sieć Bayesa

• Wydawca: Systems Research Institute, Polish Academy of Sciences
• Czasopismo: Control and Cybernetics
• Rocznik: 2000
• Tom: Vol. 29, no 4
• Strony: 921--936
• Opis fizyczny: Bibliogr. 23 poz.,

Twórcy

autor

Wierzchoń S.

autor

Kłopotek M.

• Institute of Computer Science, Polish Academy of Sciences and Dept. of Computer Science, Białystok University of Technology,