Markov Chain Monte Carlo Using Tree-Based Priors on Model Structure
Abstract
We present a general framework for defining priors on model structure and sampling from the posterior using the Metropolis-Hastings algorithm. The key ideas are that structure priors are defined via a probability tree and that the proposal distribution for the Metropolis-Hastings algorithm is defined using the prior, thereby defining a cheaply computable acceptance probability. We have applied this approach to Bayesian net structure learning using a number of priors and proposal distributions. Our results show that these must be chosen appropriately for this approach to be successful.
Cite
Text
Angelopoulos and Cussens. "Markov Chain Monte Carlo Using Tree-Based Priors on Model Structure." Conference on Uncertainty in Artificial Intelligence, 2001.Markdown
[Angelopoulos and Cussens. "Markov Chain Monte Carlo Using Tree-Based Priors on Model Structure." Conference on Uncertainty in Artificial Intelligence, 2001.](https://mlanthology.org/uai/2001/angelopoulos2001uai-markov/)BibTeX
@inproceedings{angelopoulos2001uai-markov,
title = {{Markov Chain Monte Carlo Using Tree-Based Priors on Model Structure}},
author = {Angelopoulos, Nicos and Cussens, James},
booktitle = {Conference on Uncertainty in Artificial Intelligence},
year = {2001},
pages = {16-23},
url = {https://mlanthology.org/uai/2001/angelopoulos2001uai-markov/}
}