CUED Publications database

Convolutional Gaussian Processes

van der Wilk, M and Rasmussen, CE and Hensman, J (2017) Convolutional Gaussian Processes. In: Neural Information Processing Systems 2017 (NIPS'17), 2017-12-4 to 2017-12-9, Long Beach, California, USA pp. 2845-2854..

Full text not available from this repository.


We present a practical way of introducing convolutional structure into Gaussian processes, making them more suited to high-dimensional inputs like images. The main contribution of our work is the construction of an inter-domain inducing point approximation that is well-tailored to the convolutional kernel. This allows us to gain the generalisation benefit of a convolutional kernel, together with fast but accurate posterior inference. We investigate several variations of the convolutional kernel, and apply it to MNIST and CIFAR-10, which have both been known to be challenging for Gaussian processes. We also show how the marginal likelihood can be used to find an optimal weighting between convolutional and RBF kernels to further improve performance. We hope that this illustration of the usefulness of a marginal likelihood will help automate discovering architectures in larger models.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Divisions: Div F > Computational and Biological Learning
Depositing User: Cron Job
Date Deposited: 08 Dec 2017 20:20
Last Modified: 15 Apr 2021 05:58
DOI: 10.5555/3294996.3295044