TY - GEN

T1 - Random generation of Bayesian networks

AU - Ide, Jaime S.

AU - Cozman, Fabio G.

PY - 2002/1/1

Y1 - 2002/1/1

N2 - © Springer-Verlag Berlin Heidelberg 2002. This paper presents new methods for generation of random Bayesian networks. Suchme thods can be used to test inference and learning algorithms for Bayesian networks, and to obtain insights on average properties of such networks. Any method that generates Bayesian networks must first generate directed acyclic graphs (the “structure” of the network) and then, for the generated graph, conditional probability distributions. No algorithm in the literature currently offers guarantees concerning the distribution of generated Bayesian networks. Using tools from the theory of Markov chains, we propose algorithms that can generate uniformly distributed samples of directed acyclic graphs. We introduce methods for the uniform generation of multi-connected and singly-connected networks for a given number of nodes; constraints on node degree and number of arcs can be easily imposed. After a directed acyclic graphi s uniformly generated, the conditional distributions are produced by sampling Dirichlet distributions.

AB - © Springer-Verlag Berlin Heidelberg 2002. This paper presents new methods for generation of random Bayesian networks. Suchme thods can be used to test inference and learning algorithms for Bayesian networks, and to obtain insights on average properties of such networks. Any method that generates Bayesian networks must first generate directed acyclic graphs (the “structure” of the network) and then, for the generated graph, conditional probability distributions. No algorithm in the literature currently offers guarantees concerning the distribution of generated Bayesian networks. Using tools from the theory of Markov chains, we propose algorithms that can generate uniformly distributed samples of directed acyclic graphs. We introduce methods for the uniform generation of multi-connected and singly-connected networks for a given number of nodes; constraints on node degree and number of arcs can be easily imposed. After a directed acyclic graphi s uniformly generated, the conditional distributions are produced by sampling Dirichlet distributions.

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=76349108199&origin=inward

UR - https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=76349108199&origin=inward

M3 - Conference contribution

SN - 3540001247

SN - 9783540001249

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 366

EP - 376

BT - Random generation of Bayesian networks

T2 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Y2 - 1 January 2018

ER -