Gray-box inference for structured Gaussian process models

Galliani, Pietro, Dezfouli, Amir, Bonilla, Edwin and Quadrianto, Novi (2017) Gray-box inference for structured Gaussian process models. Published in: Singh, Aarti and Zhu, Jerry, (eds.) Proceedings of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS); Fort Lauderdale, Florida, USA; 20-22 April 2017. 54 353-361. JMLR ISSN 1938-7228

[img] PDF - Published Version
Available under License Creative Commons Attribution.

Download (425kB)

Abstract

We develop an automated variational infer- ence method for Bayesian structured prediction problems with Gaussian process (gp) priors and linear-chain likelihoods. Our approach does not need to know the details of the structured likelihood model and can scale up to a large number of observations. Furthermore, we show that the required expected likelihood term and its gradients in the variational objective (ELBO) can be estimated efficiently by using expectations over very low-dimensional Gaussian distributions. Optimization of the ELBO is fully parallelizable over sequences and amenable to stochastic optimization, which we use along with control variate techniques to make our framework useful in practice. Results on a set of natural language processing tasks show that our method can be as good as (and sometimes better than, in particular with respect to expected log-likelihood) hard-coded approaches including svm-struct and crfs, and overcomes the scalability limitations of previous inference algorithms based on sampling. Overall, this is a fundamental step to developing automated inference methods for Bayesian structured prediction.

Item Type: Conference Proceedings
Schools and Departments: School of Engineering and Informatics > Informatics
Research Centres and Groups: Data Science Research Group
Subjects: Q Science > QA Mathematics > QA0276 Mathematical statistics
Related URLs:
Depositing User: Novi Quadrianto
Date Deposited: 17 Mar 2017 08:58
Last Modified: 16 Jun 2017 08:53
URI: http://sro.sussex.ac.uk/id/eprint/67122

View download statistics for this item

📧 Request an update