Black-box Importance Sampling

Qiang Liu, Jason Lee
Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, PMLR 54:952-961, 2017.

Abstract

Importance sampling is widely used in machine learning and statistics, but its power is limited by the restriction of using simple proposals for which the importance weights can be tractably calculated. We address this problem by studying black-box importance sampling methods that calculate importance weights for samples generated from any unknown proposal or black-box mechanism. Our method allows us to use better and richer proposals to solve difficult problems, and (somewhat counter-intuitively) also has the additional benefit of improving the estimation accuracy beyond typical importance sampling. Both theoretical and empirical analyses are provided.

Cite this Paper

BibTeX
@InProceedings{pmlr-v54-liu17b, title = {{Black-box Importance Sampling}}, author = {Liu, Qiang and Lee, Jason}, booktitle = {Proceedings of the 20th International Conference on Artificial Intelligence and Statistics}, pages = {952--961}, year = {2017}, editor = {Singh, Aarti and Zhu, Jerry}, volume = {54}, series = {Proceedings of Machine Learning Research}, month = {20--22 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v54/liu17b/liu17b.pdf}, url = {https://proceedings.mlr.press/v54/liu17b.html}, abstract = {Importance sampling is widely used in machine learning and statistics, but its power is limited by the restriction of using simple proposals for which the importance weights can be tractably calculated. We address this problem by studying black-box importance sampling methods that calculate importance weights for samples generated from any unknown proposal or black-box mechanism. Our method allows us to use better and richer proposals to solve difficult problems, and (somewhat counter-intuitively) also has the additional benefit of improving the estimation accuracy beyond typical importance sampling. Both theoretical and empirical analyses are provided. } }
Endnote
%0 Conference Paper %T Black-box Importance Sampling %A Qiang Liu %A Jason Lee %B Proceedings of the 20th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2017 %E Aarti Singh %E Jerry Zhu %F pmlr-v54-liu17b %I PMLR %P 952--961 %U https://proceedings.mlr.press/v54/liu17b.html %V 54 %X Importance sampling is widely used in machine learning and statistics, but its power is limited by the restriction of using simple proposals for which the importance weights can be tractably calculated. We address this problem by studying black-box importance sampling methods that calculate importance weights for samples generated from any unknown proposal or black-box mechanism. Our method allows us to use better and richer proposals to solve difficult problems, and (somewhat counter-intuitively) also has the additional benefit of improving the estimation accuracy beyond typical importance sampling. Both theoretical and empirical analyses are provided.
APA
Liu, Q. & Lee, J.. (2017). Black-box Importance Sampling. Proceedings of the 20th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 54:952-961 Available from https://proceedings.mlr.press/v54/liu17b.html.

Related Material