Bayesian Generalised Ensemble Markov Chain Monte Carlo

Jes Frellsen, Ole Winther, Zoubin Ghahramani, Jesper Ferkinghoff-Borg
Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, PMLR 51:408-416, 2016.

Abstract

Bayesian generalised ensemble (BayesGE) is a new method that addresses two major drawbacks of standard Markov chain Monte Carlo algorithms for inference in high-dimensional probability models: inapplicability to estimate the partition function and poor mixing properties. BayesGE uses a Bayesian approach to iteratively update the belief about the density of states (distribution of the log likelihood under the prior) for the model, with the dual purpose of enhancing the sampling efficiency and making the estimation of the partition function tractable. We benchmark BayesGE on Ising and Potts systems and show that it compares favourably to existing state-of-the-art methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v51-frellsen16, title = {Bayesian Generalised Ensemble Markov Chain Monte Carlo}, author = {Frellsen, Jes and Winther, Ole and Ghahramani, Zoubin and Ferkinghoff-Borg, Jesper}, booktitle = {Proceedings of the 19th International Conference on Artificial Intelligence and Statistics}, pages = {408--416}, year = {2016}, editor = {Gretton, Arthur and Robert, Christian C.}, volume = {51}, series = {Proceedings of Machine Learning Research}, address = {Cadiz, Spain}, month = {09--11 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v51/frellsen16.pdf}, url = {https://proceedings.mlr.press/v51/frellsen16.html}, abstract = {Bayesian generalised ensemble (BayesGE) is a new method that addresses two major drawbacks of standard Markov chain Monte Carlo algorithms for inference in high-dimensional probability models: inapplicability to estimate the partition function and poor mixing properties. BayesGE uses a Bayesian approach to iteratively update the belief about the density of states (distribution of the log likelihood under the prior) for the model, with the dual purpose of enhancing the sampling efficiency and making the estimation of the partition function tractable. We benchmark BayesGE on Ising and Potts systems and show that it compares favourably to existing state-of-the-art methods.} }
Endnote
%0 Conference Paper %T Bayesian Generalised Ensemble Markov Chain Monte Carlo %A Jes Frellsen %A Ole Winther %A Zoubin Ghahramani %A Jesper Ferkinghoff-Borg %B Proceedings of the 19th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2016 %E Arthur Gretton %E Christian C. Robert %F pmlr-v51-frellsen16 %I PMLR %P 408--416 %U https://proceedings.mlr.press/v51/frellsen16.html %V 51 %X Bayesian generalised ensemble (BayesGE) is a new method that addresses two major drawbacks of standard Markov chain Monte Carlo algorithms for inference in high-dimensional probability models: inapplicability to estimate the partition function and poor mixing properties. BayesGE uses a Bayesian approach to iteratively update the belief about the density of states (distribution of the log likelihood under the prior) for the model, with the dual purpose of enhancing the sampling efficiency and making the estimation of the partition function tractable. We benchmark BayesGE on Ising and Potts systems and show that it compares favourably to existing state-of-the-art methods.
RIS
TY - CPAPER TI - Bayesian Generalised Ensemble Markov Chain Monte Carlo AU - Jes Frellsen AU - Ole Winther AU - Zoubin Ghahramani AU - Jesper Ferkinghoff-Borg BT - Proceedings of the 19th International Conference on Artificial Intelligence and Statistics DA - 2016/05/02 ED - Arthur Gretton ED - Christian C. Robert ID - pmlr-v51-frellsen16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 51 SP - 408 EP - 416 L1 - http://proceedings.mlr.press/v51/frellsen16.pdf UR - https://proceedings.mlr.press/v51/frellsen16.html AB - Bayesian generalised ensemble (BayesGE) is a new method that addresses two major drawbacks of standard Markov chain Monte Carlo algorithms for inference in high-dimensional probability models: inapplicability to estimate the partition function and poor mixing properties. BayesGE uses a Bayesian approach to iteratively update the belief about the density of states (distribution of the log likelihood under the prior) for the model, with the dual purpose of enhancing the sampling efficiency and making the estimation of the partition function tractable. We benchmark BayesGE on Ising and Potts systems and show that it compares favourably to existing state-of-the-art methods. ER -
APA
Frellsen, J., Winther, O., Ghahramani, Z. & Ferkinghoff-Borg, J.. (2016). Bayesian Generalised Ensemble Markov Chain Monte Carlo. Proceedings of the 19th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 51:408-416 Available from https://proceedings.mlr.press/v51/frellsen16.html.

Related Material