AISTATS 2019poster55 citations
Deep learning with differential Gaussian process flows
Pashupati Hegde, Markus Heinonen, Harri Lähdesmäki, Samuel Kaski
Abstract
We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate excellent results as compared to deep Gaussian processes and Bayesian neural networks.
BibTeX
@InProceedings{pmlr-v89-hegde19a,
title = {Deep learning with differential Gaussian process flows},
author = {Hegde, Pashupati and Heinonen, Markus and L\"ahdesm\"aki, Harri and Kaski, Samuel},
booktitle = {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics},
pages = {1812--1821},
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/hegde19a/hegde19a.pdf},
url = {https://proceedings.mlr.press/v89/hegde19a.html},
abstract = {We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate excellent results as compared to deep Gaussian processes and Bayesian neural networks.}
}