On correlation and budget constraints in model-based bandit optimization with application to automatic machine learning

Matthew Hoffman, Bobak Shahriari, Nando Freitas
Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, PMLR 33:365-374, 2014.

Abstract

We address the problem of finding the maximizer of a nonlinear function that can only be evaluated, subject to noise, at a finite number of query locations. Further, we will assume that there is a constraint on the total number of permitted function evaluations. We introduce a Bayesian approach for this problem and show that it empirically outperforms both the existing frequentist counterpart and other Bayesian optimization methods. The Bayesian approach places emphasis on detailed modelling, including the modelling of correlations among the arms. As a result, it can perform well in situations where the number of arms is much larger than the number of allowed function evaluation, whereas the frequentist counterpart is inapplicable. This feature enables us to develop and deploy practical applications, such as automatic machine learning toolboxes. The paper presents comprehensive comparisons of the proposed approach with many Bayesian and bandit optimization techniques, the first comparison of many of these methods in the literature.

Cite this Paper

BibTeX
@InProceedings{pmlr-v33-hoffman14, title = {{On correlation and budget constraints in model-based bandit optimization with application to automatic machine learning}}, author = {Hoffman, Matthew and Shahriari, Bobak and Freitas, Nando}, booktitle = {Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics}, pages = {365--374}, 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/hoffman14.pdf}, url = {https://proceedings.mlr.press/v33/hoffman14.html}, abstract = {We address the problem of finding the maximizer of a nonlinear function that can only be evaluated, subject to noise, at a finite number of query locations. Further, we will assume that there is a constraint on the total number of permitted function evaluations. We introduce a Bayesian approach for this problem and show that it empirically outperforms both the existing frequentist counterpart and other Bayesian optimization methods. The Bayesian approach places emphasis on detailed modelling, including the modelling of correlations among the arms. As a result, it can perform well in situations where the number of arms is much larger than the number of allowed function evaluation, whereas the frequentist counterpart is inapplicable. This feature enables us to develop and deploy practical applications, such as automatic machine learning toolboxes. The paper presents comprehensive comparisons of the proposed approach with many Bayesian and bandit optimization techniques, the first comparison of many of these methods in the literature.} }
Endnote
%0 Conference Paper %T On correlation and budget constraints in model-based bandit optimization with application to automatic machine learning %A Matthew Hoffman %A Bobak Shahriari %A Nando Freitas %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-hoffman14 %I PMLR %P 365--374 %U https://proceedings.mlr.press/v33/hoffman14.html %V 33 %X We address the problem of finding the maximizer of a nonlinear function that can only be evaluated, subject to noise, at a finite number of query locations. Further, we will assume that there is a constraint on the total number of permitted function evaluations. We introduce a Bayesian approach for this problem and show that it empirically outperforms both the existing frequentist counterpart and other Bayesian optimization methods. The Bayesian approach places emphasis on detailed modelling, including the modelling of correlations among the arms. As a result, it can perform well in situations where the number of arms is much larger than the number of allowed function evaluation, whereas the frequentist counterpart is inapplicable. This feature enables us to develop and deploy practical applications, such as automatic machine learning toolboxes. The paper presents comprehensive comparisons of the proposed approach with many Bayesian and bandit optimization techniques, the first comparison of many of these methods in the literature.
RIS
TY - CPAPER TI - On correlation and budget constraints in model-based bandit optimization with application to automatic machine learning AU - Matthew Hoffman AU - Bobak Shahriari AU - Nando Freitas 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-hoffman14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 33 SP - 365 EP - 374 L1 - http://proceedings.mlr.press/v33/hoffman14.pdf UR - https://proceedings.mlr.press/v33/hoffman14.html AB - We address the problem of finding the maximizer of a nonlinear function that can only be evaluated, subject to noise, at a finite number of query locations. Further, we will assume that there is a constraint on the total number of permitted function evaluations. We introduce a Bayesian approach for this problem and show that it empirically outperforms both the existing frequentist counterpart and other Bayesian optimization methods. The Bayesian approach places emphasis on detailed modelling, including the modelling of correlations among the arms. As a result, it can perform well in situations where the number of arms is much larger than the number of allowed function evaluation, whereas the frequentist counterpart is inapplicable. This feature enables us to develop and deploy practical applications, such as automatic machine learning toolboxes. The paper presents comprehensive comparisons of the proposed approach with many Bayesian and bandit optimization techniques, the first comparison of many of these methods in the literature. ER -
APA
Hoffman, M., Shahriari, B. & Freitas, N.. (2014). On correlation and budget constraints in model-based bandit optimization with application to automatic machine learning. Proceedings of the Seventeenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 33:365-374 Available from https://proceedings.mlr.press/v33/hoffman14.html.

Related Material