Wyniki wyszukiwania - BazTech

1

Reasoning methods in general and stuctured Bayesian networks

Kłopotek M. A.

Studia Informatica : systems and information technology

|

2003

|

Vol. 1(1)

5--25

EN

Bayesian networks have many practical applications due to their capability to represent joint probability distribution in many variables in a compact way. Though there exist many algorithms for learning Bayesian networks from data, they are not satisfactory because the learned networks usually are not suitable for reasoning. So far only a restricted class of very simple Bayesian networks: trees and poly-trees are directly applicable in reasoning. This paper defines and explores a new class of networks: the Structured Bayesian Networks. Two methods of reasoning are outlined for this type of networks. Possible methods of learning from data are indicated. Similarity to hiearachical networks is pointed at.

2

Reasoning and Learning in Extended Structured Bayesian Networks

Kłopotek M. A.

Fundamenta Informaticae

|

2003

|

Vol. 58, nr 2

105--137

EN

Bayesian networks have many practical applications due to their capability to represent joint probability distribution over many variables in a compact way. Though there exist many algorithms for learning Bayesian networks from data, they are not satisfactory because the learned networks usually are not suitable directly for reasoning as they need to be transformed to some other form (tree, polytree, hypertree) statically or dynamically, and this transformation is not trivial [25]. So far only a restricted class of very simple Bayesian networks: trees and poly-trees are directly applicable in reasoning. This paper defines and explores a new class of networks: the Structured Bayesian Networks. Two methods of reasoning are outlined for this type of networks. Possible methods of learning from data are indicated. Similarity to hierarchical networks is pointed at.

3

Structure and Reasoning in Bayesian Networks

Kłopotek M.

Prace Instytutu Podstaw Informatyki Polskiej Akademii Nauk

|

2001

|

Nr 941

1-12

EN

Bayesian networks have many practical applications due to their capability to represent joint probability distribution in many variables in a compact way. Though there exist many algorithms for learning Bayesian networks from data, they are not satisfactory because the learned networks usually are not suitable for reasoning. So far only a restricted class of very simple Bayesian networks: trees and poly-trees are directly applicable in reasoning. This paper defines and explores a new class of networks: the Structured Bayesian Networks. Two methods of reasoning are outlined for this type of networks. Possible methods of learning from data are indicated. Similarity to hierarchical networks is pointed at.

PL

Sieci bayesowskie maja wiele praktycznych zastosowań związanych z ich zdolnością do zwartej reprezentacji rozkładów prawdopodobieństwa w wielu zmiennych. Choć znanych jest wiele algorytmów uczących sieci bayesowskie z danych, nie są one satysfakcjonujące, ponieważ wynikowe struktury sieci na ogół nie nadają się do celów wnioskowania eksperckiego, z wyjątkiem bardzo prostych struktur drzew i polidrzew. Niniejszy artykuł definiuje i bada nową klasę sieci bayesowskich: tzw. strukturalne sieci bayesowskie. Przedstawiono dwie metody wnioskowania dla tych sieci. Wskazano na możliwość uczenia się tych sieci z danych. Sieci te można uważać za interesujący szczególny przypadek sieci hierarchicznych.