Kernel Conditional Exponential Family

Michael Arbel, Arthur Gretton
Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR 84:1337-1346, 2018.

Abstract

A nonparametric family of conditional distributions is introduced, which generalizes conditional exponential families using functional parameters in a suitable RKHS. An algorithm is provided for learning the generalized natural parameter, and consistency of the estimator is established in the well specified case. In experiments, the new method generally outperforms a competing approach with consistency guarantees, and is competitive with a deep conditional density model on datasets that exhibit abrupt transitions and heteroscedasticity.

Cite this Paper


BibTeX
@InProceedings{pmlr-v84-arbel18a, title = {Kernel Conditional Exponential Family}, author = {Arbel, Michael and Gretton, Arthur}, booktitle = {Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics}, pages = {1337--1346}, year = {2018}, editor = {Storkey, Amos and Perez-Cruz, Fernando}, volume = {84}, series = {Proceedings of Machine Learning Research}, month = {09--11 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v84/arbel18a/arbel18a.pdf}, url = {https://proceedings.mlr.press/v84/arbel18a.html}, abstract = {A nonparametric family of conditional distributions is introduced, which generalizes conditional exponential families using functional parameters in a suitable RKHS. An algorithm is provided for learning the generalized natural parameter, and consistency of the estimator is established in the well specified case. In experiments, the new method generally outperforms a competing approach with consistency guarantees, and is competitive with a deep conditional density model on datasets that exhibit abrupt transitions and heteroscedasticity. } }
Endnote
%0 Conference Paper %T Kernel Conditional Exponential Family %A Michael Arbel %A Arthur Gretton %B Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2018 %E Amos Storkey %E Fernando Perez-Cruz %F pmlr-v84-arbel18a %I PMLR %P 1337--1346 %U https://proceedings.mlr.press/v84/arbel18a.html %V 84 %X A nonparametric family of conditional distributions is introduced, which generalizes conditional exponential families using functional parameters in a suitable RKHS. An algorithm is provided for learning the generalized natural parameter, and consistency of the estimator is established in the well specified case. In experiments, the new method generally outperforms a competing approach with consistency guarantees, and is competitive with a deep conditional density model on datasets that exhibit abrupt transitions and heteroscedasticity.
APA
Arbel, M. & Gretton, A.. (2018). Kernel Conditional Exponential Family. Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 84:1337-1346 Available from https://proceedings.mlr.press/v84/arbel18a.html.

Related Material