Black Box Quantiles for Kernel Learning

Anthony Tompkins, Ransalu Senanayake, Philippe Morere, Fabio Ramos
Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, PMLR 89:1427-1437, 2019.

Abstract

Kernel methods have been successfully used in various domains to model nonlinear patterns. However, the structure of the kernels is typically handcrafted for each dataset based on the experience of the data analyst. In this paper, we present a novel technique to learn kernels that best fit the data. We exploit the measure-theoretic view of a shift-invariant kernel given by the Bochner’s theorem, and automatically learn the measure in terms of a parameterized quantile function. This flexible black box quantile function, evaluated on Quasi-Monte Carlo samples, builds up quasi-random Fourier feature maps that can approximate arbitrary kernels. The proposed method is not only general enough to be used in any kernel machine, but can also be combined with other kernel design techniques. We learn expressive kernels on a variety of datasets, verifying the methods ability to automatically discover complex patterns without being guided by human expert knowledge.

Cite this Paper


BibTeX
@InProceedings{pmlr-v89-tompkins19a, title = {Black Box Quantiles for Kernel Learning}, author = {Tompkins, Anthony and Senanayake, Ransalu and Morere, Philippe and Ramos, Fabio}, booktitle = {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics}, pages = {1427--1437}, year = {2019}, editor = {Chaudhuri, Kamalika and Sugiyama, Masashi}, volume = {89}, series = {Proceedings of Machine Learning Research}, month = {16--18 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v89/tompkins19a/tompkins19a.pdf}, url = {https://proceedings.mlr.press/v89/tompkins19a.html}, abstract = {Kernel methods have been successfully used in various domains to model nonlinear patterns. However, the structure of the kernels is typically handcrafted for each dataset based on the experience of the data analyst. In this paper, we present a novel technique to learn kernels that best fit the data. We exploit the measure-theoretic view of a shift-invariant kernel given by the Bochner’s theorem, and automatically learn the measure in terms of a parameterized quantile function. This flexible black box quantile function, evaluated on Quasi-Monte Carlo samples, builds up quasi-random Fourier feature maps that can approximate arbitrary kernels. The proposed method is not only general enough to be used in any kernel machine, but can also be combined with other kernel design techniques. We learn expressive kernels on a variety of datasets, verifying the methods ability to automatically discover complex patterns without being guided by human expert knowledge.} }
Endnote
%0 Conference Paper %T Black Box Quantiles for Kernel Learning %A Anthony Tompkins %A Ransalu Senanayake %A Philippe Morere %A Fabio Ramos %B Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Masashi Sugiyama %F pmlr-v89-tompkins19a %I PMLR %P 1427--1437 %U https://proceedings.mlr.press/v89/tompkins19a.html %V 89 %X Kernel methods have been successfully used in various domains to model nonlinear patterns. However, the structure of the kernels is typically handcrafted for each dataset based on the experience of the data analyst. In this paper, we present a novel technique to learn kernels that best fit the data. We exploit the measure-theoretic view of a shift-invariant kernel given by the Bochner’s theorem, and automatically learn the measure in terms of a parameterized quantile function. This flexible black box quantile function, evaluated on Quasi-Monte Carlo samples, builds up quasi-random Fourier feature maps that can approximate arbitrary kernels. The proposed method is not only general enough to be used in any kernel machine, but can also be combined with other kernel design techniques. We learn expressive kernels on a variety of datasets, verifying the methods ability to automatically discover complex patterns without being guided by human expert knowledge.
APA
Tompkins, A., Senanayake, R., Morere, P. & Ramos, F.. (2019). Black Box Quantiles for Kernel Learning. Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 89:1427-1437 Available from https://proceedings.mlr.press/v89/tompkins19a.html.

Related Material