Multi-task Gaussian Process Prediction

Multi-task Gaussian Process Prediction

Cached

Download Links

by Edwin V. Bonilla , Kian Ming , A. Chai , Christopher K. I. Williams

Citations:

BibTeX

@MISC{Bonilla_multi-taskgaussian,
    author = {Edwin V. Bonilla and Kian Ming and A. Chai and Christopher K. I. Williams},
    title = {Multi-task Gaussian Process Prediction},
    year = {}
}

Bookmark

OpenURL

Abstract

In this paper we investigate multi-task learning in the context of Gaussian Processes (GP). We propose a model that learns a shared covariance function on input-dependent features and a “free-form ” covariance matrix over tasks. This allows for good flexibility when modelling inter-task dependencies while avoiding the need for large amounts of data for training. We show that under the assumption of noise-free observations and a block design, predictions for a given task only depend on its target values and therefore a cancellation of inter-task transfer occurs. We evaluate the benefits of our model on two practical applications: a compiler performance prediction problem and an exam score prediction task. Additionally, we make use of GP approximations and properties of our model in order to provide scalability to large data sets. 1

Citations

359	Probabilistic principal component analysis - Tipping, Bishop - 1999
109	Is learning the n-th thing any easier than learning the first - Thrun - 1996
100	A model of inductive bias learning - Baxter - 2000
96	Learning multiple tasks with kernel methods - Evgeniou - 2005
74	Learning gaussian processes from multiple tasks - Yu, Tresp, et al. - 2005
72	Task clustering and gating for bayesian multitask learning - Bakker, Heskes - 2003
42	Multivariate geostatistics : an introduction with applications - Wackernagel - 2003
37	Learning to learn with the informative vector machine - Lawrence, Platt - 2004
34	Semiparametric latent factor models - Teh, Seeger - 2005
19	X.: Stochastic Relational Models for Discriminative Link Prediction - Yu, Chu, et al. - 2007
19	Learning Gaussian process kernels via hierarchical Bayes - Schwaighofer, Tresp, et al.
13	Kernel multi-task learning using task-specific features - Bonilla, Agakov, et al. - 2007
11	Collaborative ordinal regression - Yu, Yu, et al. - 2006
9	Approximation methods for Gaussian process regression - Quinonero-Candela, Rasmussen, et al. - 2007
6	Learning how to learn is learning with point sets - Minka, Picard - 1997
3	Maximum-likelihood estimation for multivariate spatial linear coregionalization models. Environmetrics - Zhang - 2007
2	A note on noise-free Gaussian process prediction with separable covariance functions and grid designs. Informatics Research Report 1228 - I, Chai, et al. - 2007
1	A Markov property for covariance structures - O’Hagan - 1998