Composing inference algorithms as program transformations

Robert Zinkov, Chung-chieh Shan
Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, PMLR R15:191-200, 2017.

Abstract

Probabilistic inference procedures are usually coded painstakingly from scratch, for each target model and each inference algorithm. We reduce this effort by generating inference procedures from models automatically. We make this code generation modular by decomposing inference algorithms into reusable program-to program transformations. These transformations perform exact inference as well as generate probabilistic programs that compute expectations, densities, and MCMC samples. The resulting inference procedures are about as accurate and fast as other probabilistic programming systems on real-world problems.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR15-zinkov17a, title = {Composing inference algorithms as program transformations}, author = {Zinkov, Robert and Shan, Chung-chieh}, booktitle = {Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence}, pages = {191--200}, year = {2017}, editor = {Elidan, Gal and Kersting, Kristian}, volume = {R15}, series = {Proceedings of Machine Learning Research}, month = {11--15 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r15/main/assets/zinkov17a/zinkov17a.pdf}, url = {https://proceedings.mlr.press/r15/zinkov17a.html}, abstract = {Probabilistic inference procedures are usually coded painstakingly from scratch, for each target model and each inference algorithm. We reduce this effort by generating inference procedures from models automatically. We make this code generation modular by decomposing inference algorithms into reusable program-to program transformations. These transformations perform exact inference as well as generate probabilistic programs that compute expectations, densities, and MCMC samples. The resulting inference procedures are about as accurate and fast as other probabilistic programming systems on real-world problems.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Composing inference algorithms as program transformations %A Robert Zinkov %A Chung-chieh Shan %B Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2017 %E Gal Elidan %E Kristian Kersting %F pmlr-vR15-zinkov17a %I PMLR %P 191--200 %U https://proceedings.mlr.press/r15/zinkov17a.html %V R15 %X Probabilistic inference procedures are usually coded painstakingly from scratch, for each target model and each inference algorithm. We reduce this effort by generating inference procedures from models automatically. We make this code generation modular by decomposing inference algorithms into reusable program-to program transformations. These transformations perform exact inference as well as generate probabilistic programs that compute expectations, densities, and MCMC samples. The resulting inference procedures are about as accurate and fast as other probabilistic programming systems on real-world problems. %Z Reissued by PMLR on 04 October 2026.
APA
Zinkov, R. & Shan, C.. (2017). Composing inference algorithms as program transformations. Proceedings of the 33rd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R15:191-200 Available from https://proceedings.mlr.press/r15/zinkov17a.html. Reissued by PMLR on 04 October 2026.

Related Material