Decision Sum-Product-Max Networks

Mazen Melibari, Pascal Poupart, Prashant Doshi

AAAI 2016 pp. 4234-4235

doi:10.1609/AAAI.V30I1.9957 /aaai/2016/melibari2016aaai-decision/

Abstract

Sum-Product Networks (SPNs) were recently proposed as a new class of probabilistic graphical models that guarantee tractable inference, even on models with high-treewidth. In this paper, we propose a new extension to SPNs, called Decision Sum-Product-Max Networks (Decision-SPMNs), that makes SPNs suitable for discrete multi-stage decision problems. We present an algorithm that solves Decision-SPMNs in a time that is linear in the size of the network. We also present algorithms to learn the parameters of the network from data.

PDF AAAI Semantic Scholar

Cite

Text

Melibari et al. "Decision Sum-Product-Max Networks." AAAI Conference on Artificial Intelligence, 2016. doi:10.1609/AAAI.V30I1.9957

Markdown

[Melibari et al. "Decision Sum-Product-Max Networks." AAAI Conference on Artificial Intelligence, 2016.](https://mlanthology.org/aaai/2016/melibari2016aaai-decision/) doi:10.1609/AAAI.V30I1.9957

BibTeX

@inproceedings{melibari2016aaai-decision,
  title     = {{Decision Sum-Product-Max Networks}},
  author    = {Melibari, Mazen and Poupart, Pascal and Doshi, Prashant},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2016},
  pages     = {4234-4235},
  doi       = {10.1609/AAAI.V30I1.9957},
  url       = {https://mlanthology.org/aaai/2016/melibari2016aaai-decision/}
}