Calibrating deep convolutional Gaussian processes

G. L. Tran, E. V. Bonilla, J. P. Cunningham, P. Michiardi, M. Filippone

Research output: Contribution to conferencePaperpeer-review

13 Scopus citations

Abstract

The wide adoption of Convolutional Neural Networks (cnns) in applications where decision-making under uncertainty is fundamental, has brought a great deal of attention to the ability of these models to accurately quantify the uncertainty in their predictions. Previous work on combining cnns with Gaussian processes (gps) has been developed under the assumption that the predictive probabilities of these models are well-calibrated. In this paper we show that, in fact, current combinations of cnns and gps are miscalibrated. We propose a novel combination that considerably outperforms previous approaches on this aspect, while achieving state-of-the-art performance on image classification tasks.

Original languageEnglish (US)
StatePublished - 2020
Event22nd International Conference on Artificial Intelligence and Statistics, AISTATS 2019 - Naha, Japan
Duration: Apr 16 2019Apr 18 2019

Conference

Conference22nd International Conference on Artificial Intelligence and Statistics, AISTATS 2019
Country/TerritoryJapan
CityNaha
Period04/16/1904/18/19

Bibliographical note

Publisher Copyright:
© Copyright 2019 by the author(s).

ASJC Scopus subject areas

  • Artificial Intelligence
  • Statistics and Probability

Fingerprint

Dive into the research topics of 'Calibrating deep convolutional Gaussian processes'. Together they form a unique fingerprint.

Cite this