Conference item icon

Conference item

BRUNO: A deep recurrent model for exchangeable data

Abstract:
We present a novel model architecture which leverages deep learning tools to perform exact Bayesian inference on sets of high dimensional, complex observations. Our model is provably exchangeable, meaning that the joint distribution over observations is invariant under permutation: this property lies at the heart of Bayesian inference. The model does not require variational approximations to train, and new samples can be generated conditional on previous samples, with cost linear in the size of the conditioning set. The advantages of our architecture are demonstrated on learning tasks that require generalisation from short observed sequences while modelling sequence variability, such as conditional image generation, few-shot learning, and anomaly detection.
Publication status:
Published
Peer review status:
Peer reviewed

Actions


Access Document


Authors


More by this author
Institution:
University of Oxford
Division:
MPLS
Department:
Computer Science
Oxford college:
Christ Church
Role:
Author


Publisher:
Massachusetts Institute of Technology Press
Host title:
Advances in Neural Information Processing Systems
Journal:
Advances in Neural Information Processing Systems More from this journal
Publication date:
2018-12-01
Acceptance date:
2018-09-05
Event location:
Montreal
ISSN:
1049-5258


Pubs id:
pubs:1045797
UUID:
uuid:b742c28e-9f2a-4b5a-926c-f512306fa5c6
Local pid:
pubs:1045797
Source identifiers:
1045797
Deposit date:
2019-08-16

Terms of use



Views and Downloads






If you are the owner of this record, you can report an update to it here: Report update to this record

TO TOP