Rethinking Variational Inference for Probabilistic Programs with Stochastic Support

Abstract

We introduce Support Decomposition Variational Inference (SDVI), a new variational inference (VI) approach for probabilistic programs with stochastic support. Existing approaches to this problem rely on designing a single global variational guide on a variable-by-variable basis, while maintaining the stochastic control flow of the original program. SDVI instead breaks the program down into sub-programs with static support, before automatically building separate sub-guides for each. This decomposition significantly aids in the construction of suitable variational families, enabling, in turn, substantial improvements in inference performance.

PDF NeurIPS OpenReview Semantic Scholar

Cite

Text

Reichelt et al. "Rethinking Variational Inference for Probabilistic Programs with Stochastic Support." Neural Information Processing Systems, 2022.

Markdown

[Reichelt et al. "Rethinking Variational Inference for Probabilistic Programs with Stochastic Support." Neural Information Processing Systems, 2022.](https://mlanthology.org/neurips/2022/reichelt2022neurips-rethinking/)

BibTeX

@inproceedings{reichelt2022neurips-rethinking,
  title     = {{Rethinking Variational Inference for Probabilistic Programs with Stochastic Support}},
  author    = {Reichelt, Tim and Ong, Luke and Rainforth, Thomas},
  booktitle = {Neural Information Processing Systems},
  year      = {2022},
  url       = {https://mlanthology.org/neurips/2022/reichelt2022neurips-rethinking/}
}