Gaussian Process Regression Networks

Abstract

We introduce a new regression framework, Gaussian process regression networks (GPRN), which combines the structural properties of Bayesian neural networks with the non-parametric flexibility of Gaussian processes. This model accommodates input dependent signal and noise correlations between multiple response variables, input dependent length-scales and amplitudes, and heavy-tailed predictive distributions. We derive both efficient Markov chain Monte Carlo and variational Bayes inference procedures for this model. We apply GPRN as a multiple output regression and multivariate volatility model, demonstrating substantially improved performance over eight popular multiple output (multi-task) Gaussian process models and three multivariate volatility models on benchmark datasets, including a 1000 dimensional gene expression dataset.

17 pages, 3 figures, 1 table. Submitted for publication

References in corpus (1)

• Gaussian Process Networks

Cited by in corpus (25)

• Scalable Variational Gaussian Process Classification
• Gaussian Process Kernels for Pattern Discovery and Extrapolation
• Deep Gaussian Processes
• 2022 Review of Data-Driven Plasma Science
• Active emulation of computer codes with Gaussian processes -- Application to remote sensing
• Hidden Parameter Markov Decision Processes: A Semiparametric Regression Approach for Discovering Latent Task Parametrizations
• Exact Gaussian Processes on a Million Data Points
• Multi-resolution Multi-task Gaussian Processes
• Deep Kernel Learning
• The Gaussian Process Autoregressive Regression Model (GPAR)
• Scalable Exact Inference in Multi-Output Gaussian Processes
• Deep Multi-Fidelity Active Learning of High-dimensional Outputs
• Physiological Gaussian Process Priors for the Hemodynamics in fMRI Analysis
• Multi-task Learning for Aggregated Data using Gaussian Processes
• Variational zero-inflated Gaussian processes with sparse kernels
• Sparse Spectrum Warped Input Measures for Nonstationary Kernel Learning
• Multi-Fidelity High-Order Gaussian Processes for Physical Simulation
• Deep Coregionalization for the Emulation of Spatial-Temporal Fields
• Change Surfaces for Expressive Multidimensional Changepoints and Counterfactual Prediction
• Deconditional Downscaling with Gaussian Processes
• End-to-End Learning of Deep Kernel Acquisition Functions for Bayesian Optimization
• Environmental Modeling Framework using Stacked Gaussian Processes
• On the Constrained Least-cost Tour Problem
• Scalable Grouped Gaussian Processes via Direct Cholesky Functional Representations
• A Variant of Gaussian Process Dynamical Systems