Approximate Slice Sampling for Bayesian Posterior Inference

Christopher DuBois, Anoop Korattikara, Max Welling, Padhraic Smyth
Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, PMLR 33:185-193, 2014.

Abstract

In this paper, we advance the theory of large scale Bayesian posterior inference by introducing a new approximate slice sampler that uses only small mini-batches of data in every iteration. While this introduces a bias in the stationary distribution, the computational savings allow us to draw more samples in a given amount of time and reduce sampling variance. We empirically verify on three different models that the approximate slice sampling algorithm can significantly outperform a traditional slice sampler if we are allowed only a fixed amount of computing time for our simulations.

Cite this Paper


BibTeX
@InProceedings{pmlr-v33-dubois14, title = {{Approximate Slice Sampling for Bayesian Posterior Inference}}, author = {DuBois, Christopher and Korattikara, Anoop and Welling, Max and Smyth, Padhraic}, booktitle = {Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics}, pages = {185--193}, year = {2014}, editor = {Kaski, Samuel and Corander, Jukka}, volume = {33}, series = {Proceedings of Machine Learning Research}, address = {Reykjavik, Iceland}, month = {22--25 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v33/dubois14.pdf}, url = {https://proceedings.mlr.press/v33/dubois14.html}, abstract = {In this paper, we advance the theory of large scale Bayesian posterior inference by introducing a new approximate slice sampler that uses only small mini-batches of data in every iteration. While this introduces a bias in the stationary distribution, the computational savings allow us to draw more samples in a given amount of time and reduce sampling variance. We empirically verify on three different models that the approximate slice sampling algorithm can significantly outperform a traditional slice sampler if we are allowed only a fixed amount of computing time for our simulations.} }
Endnote
%0 Conference Paper %T Approximate Slice Sampling for Bayesian Posterior Inference %A Christopher DuBois %A Anoop Korattikara %A Max Welling %A Padhraic Smyth %B Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2014 %E Samuel Kaski %E Jukka Corander %F pmlr-v33-dubois14 %I PMLR %P 185--193 %U https://proceedings.mlr.press/v33/dubois14.html %V 33 %X In this paper, we advance the theory of large scale Bayesian posterior inference by introducing a new approximate slice sampler that uses only small mini-batches of data in every iteration. While this introduces a bias in the stationary distribution, the computational savings allow us to draw more samples in a given amount of time and reduce sampling variance. We empirically verify on three different models that the approximate slice sampling algorithm can significantly outperform a traditional slice sampler if we are allowed only a fixed amount of computing time for our simulations.
RIS
TY - CPAPER TI - Approximate Slice Sampling for Bayesian Posterior Inference AU - Christopher DuBois AU - Anoop Korattikara AU - Max Welling AU - Padhraic Smyth BT - Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics DA - 2014/04/02 ED - Samuel Kaski ED - Jukka Corander ID - pmlr-v33-dubois14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 33 SP - 185 EP - 193 L1 - http://proceedings.mlr.press/v33/dubois14.pdf UR - https://proceedings.mlr.press/v33/dubois14.html AB - In this paper, we advance the theory of large scale Bayesian posterior inference by introducing a new approximate slice sampler that uses only small mini-batches of data in every iteration. While this introduces a bias in the stationary distribution, the computational savings allow us to draw more samples in a given amount of time and reduce sampling variance. We empirically verify on three different models that the approximate slice sampling algorithm can significantly outperform a traditional slice sampler if we are allowed only a fixed amount of computing time for our simulations. ER -
APA
DuBois, C., Korattikara, A., Welling, M. & Smyth, P.. (2014). Approximate Slice Sampling for Bayesian Posterior Inference. Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 33:185-193 Available from https://proceedings.mlr.press/v33/dubois14.html.

Related Material