Approximate Structure Learning for Large Bayesian Networks

Abstract

We present approximate structure learning algorithms for Bayesian networks. We discuss the two main phases of the task: the preparation of the cache of the scores and structure optimization, both with bounded and unbounded treewidth. We improve on state-of-the-art methods that rely on an ordering-based search by sampling more effectively the space of the orders. This allows for a remarkable improvement in learning Bayesian networks from thousands of variables. We also present a thorough study of the accuracy and the running time of inference, comparing bounded-treewidth and unbounded-treewidth models.

Cite

Text

Scanagatta et al. "Approximate Structure Learning for Large Bayesian Networks." Machine Learning, 2018. doi:10.1007/S10994-018-5701-9

Markdown

[Scanagatta et al. "Approximate Structure Learning for Large Bayesian Networks." Machine Learning, 2018.](https://mlanthology.org/mlj/2018/scanagatta2018mlj-approximate/) doi:10.1007/S10994-018-5701-9

BibTeX

@article{scanagatta2018mlj-approximate,
  title     = {{Approximate Structure Learning for Large Bayesian Networks}},
  author    = {Scanagatta, Mauro and Corani, Giorgio and de Campos, Cassio Polpo and Zaffalon, Marco},
  journal   = {Machine Learning},
  year      = {2018},
  pages     = {1209-1227},
  doi       = {10.1007/S10994-018-5701-9},
  volume    = {107},
  url       = {https://mlanthology.org/mlj/2018/scanagatta2018mlj-approximate/}
}