ML Anthology

Abstraction Sampling in Graphical Models

Filjor Broka, Rina Dechter, Alexander Ihler, Kalev Kask
UAI 2018 pp. 632-641
doi:10.1609/aaai.v32i1.11365 /uai/2018/broka2018uai-abstraction/

Abstract

We present a new sampling scheme for approximating hard to compute queries over graphical models, such as computing the partition function. The scheme builds upon exact algorithms that traverse a weighted directed state-space graph representing a global function over a graphical model (e.g., probability distribution). With the aid of an abstraction function and randomization, the state space can be compacted (trimmed) to facilitate tractable computation, yielding a Monte Carlo estimate that is unbiased. We present the general idea and analyze its properties analytically and empirically.

PDF UAI Semantic Scholar

Cite

Text

Broka et al. "Abstraction Sampling in Graphical Models." Conference on Uncertainty in Artificial Intelligence, 2018. doi:10.1609/aaai.v32i1.11365

Markdown

[Broka et al. "Abstraction Sampling in Graphical Models." Conference on Uncertainty in Artificial Intelligence, 2018.](https://mlanthology.org/uai/2018/broka2018uai-abstraction/) doi:10.1609/aaai.v32i1.11365

BibTeX

@inproceedings{broka2018uai-abstraction,
  title     = {{Abstraction Sampling in Graphical Models}},
  author    = {Broka, Filjor and Dechter, Rina and Ihler, Alexander and Kask, Kalev},
  booktitle = {Conference on Uncertainty in Artificial Intelligence},
  year      = {2018},
  pages     = {632-641},
  doi       = {10.1609/aaai.v32i1.11365},
  url       = {https://mlanthology.org/uai/2018/broka2018uai-abstraction/}
}