BOCK : Bayesian Optimization with Cylindrical Kernels

ChangYong Oh, Efstratios Gavves, Max Welling
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:3868-3877, 2018.

Abstract

A major challenge in Bayesian Optimization is the boundary issue where an algorithm spends too many evaluations near the boundary of its search space. In this paper, we propose BOCK, Bayesian Optimization with Cylindrical Kernels, whose basic idea is to transform the ball geometry of the search space using a cylindrical transformation. Because of the transformed geometry, the Gaussian Process-based surrogate model spends less budget searching near the boundary, while concentrating its efforts relatively more near the center of the search region, where we expect the solution to be located. We evaluate BOCK extensively, showing that it is not only more accurate and efficient, but it also scales successfully to problems with a dimensionality as high as 500. We show that the better accuracy and scalability of BOCK even allows optimizing modestly sized neural network layers, as well as neural network hyperparameters.

Cite this Paper

BibTeX
@InProceedings{pmlr-v80-oh18a, title = {{BOCK} : {B}ayesian Optimization with Cylindrical Kernels}, author = {Oh, ChangYong and Gavves, Efstratios and Welling, Max}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {3868--3877}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/oh18a/oh18a.pdf}, url = {https://proceedings.mlr.press/v80/oh18a.html}, abstract = {A major challenge in Bayesian Optimization is the boundary issue where an algorithm spends too many evaluations near the boundary of its search space. In this paper, we propose BOCK, Bayesian Optimization with Cylindrical Kernels, whose basic idea is to transform the ball geometry of the search space using a cylindrical transformation. Because of the transformed geometry, the Gaussian Process-based surrogate model spends less budget searching near the boundary, while concentrating its efforts relatively more near the center of the search region, where we expect the solution to be located. We evaluate BOCK extensively, showing that it is not only more accurate and efficient, but it also scales successfully to problems with a dimensionality as high as 500. We show that the better accuracy and scalability of BOCK even allows optimizing modestly sized neural network layers, as well as neural network hyperparameters.} }
Endnote
%0 Conference Paper %T BOCK : Bayesian Optimization with Cylindrical Kernels %A ChangYong Oh %A Efstratios Gavves %A Max Welling %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-oh18a %I PMLR %P 3868--3877 %U https://proceedings.mlr.press/v80/oh18a.html %V 80 %X A major challenge in Bayesian Optimization is the boundary issue where an algorithm spends too many evaluations near the boundary of its search space. In this paper, we propose BOCK, Bayesian Optimization with Cylindrical Kernels, whose basic idea is to transform the ball geometry of the search space using a cylindrical transformation. Because of the transformed geometry, the Gaussian Process-based surrogate model spends less budget searching near the boundary, while concentrating its efforts relatively more near the center of the search region, where we expect the solution to be located. We evaluate BOCK extensively, showing that it is not only more accurate and efficient, but it also scales successfully to problems with a dimensionality as high as 500. We show that the better accuracy and scalability of BOCK even allows optimizing modestly sized neural network layers, as well as neural network hyperparameters.
APA
Oh, C., Gavves, E. & Welling, M.. (2018). BOCK : Bayesian Optimization with Cylindrical Kernels. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:3868-3877 Available from https://proceedings.mlr.press/v80/oh18a.html.

Related Material