Subset-of-Data Variational Inference for Deep Gaussian-Processes Regression

Jain, A. and Srijith, P K and Khan, M.E. (2021) Subset-of-Data Variational Inference for Deep Gaussian-Processes Regression. In: 37th Conference on Uncertainty in Artificial Intelligence, UAI 2021, 27 July 2021through 30 July 2021, Virtual, Online.

Full text not available from this repository. (Request a copy)

Abstract

Deep Gaussian Processes (DGPs) are multi-layer, flexible extensions of Gaussian Processes but their training remains challenging. Sparse approximations simplify the training but often require optimization over a large number of inducing inputs and their locations across layers. In this paper, we simplify the training by setting the locations to a fixed subset of data and sampling the inducing inputs from a variational distribution. This reduces the trainable parameters and computation cost without significant performance degradations, as demonstrated by our empirical results on regression problems. Our modifications simplify and stabilize DGP training while making it amenable to sampling schemes for setting the inducing inputs. © 2021 37th Conference on Uncertainty in Artificial Intelligence, UAI 2021. All Rights Reserved.

[error in script]
IITH Creators:
IITH CreatorsORCiD
Srijith, P Khttps://orcid.org/0000-0002-2820-0835
Item Type: Conference or Workshop Item (Paper)
Additional Information: MEK would like to thank a number of people who, over the years, spent time in checking the validity of (7), including Wu Lin, Heiko Strathman, Didrik Nielsen, Si Kai Lee, Anand Subramanian, Paul Cheng, and Arno Solin. AJ and PKS thank the funding and travel support from Science and Engineering Research Board (SERB), India and Japan International Co-operation Agency (JICA), Japan.
Uncontrolled Keywords: Computation costs; Gaussian process regression; Gaussian Processes; Multi-layers; Optimisations; Performance degradation; Regression problem; Sampling schemes; Sparse approximations; Variational inference
Subjects: Computer science
Divisions: Department of Computer Science & Engineering
Depositing User: . LibTrainee 2021
Date Deposited: 01 Oct 2022 05:32
Last Modified: 01 Oct 2022 05:32
URI: http://raiith.iith.ac.in/id/eprint/10756
Publisher URL:
Related URLs:

Actions (login required)

View Item View Item
Statistics for RAIITH ePrint 10756 Statistics for this ePrint Item