A Framework for Interdomain and Multioutput Gaussian Processes
arXiv:2003.01115
Abstract
One obstacle to the use of Gaussian processes (GPs) in large-scale problems, and as a component in deep learning system, is the need for bespoke derivations and implementations for small variations in the model or inference. In order to improve the utility of GPs we need a modular system that allows rapid implementation and testing, as seen in the neural network community. We present a mathematical and software framework for scalable approximate inference in GPs, which combines interdomain approximations and multiple outputs. Our framework, implemented in GPflow, provides a unified interface for many existing multioutput models, as well as more recent convolutional structures. This simplifies the creation of deep models with GPs, and we hope that this work will encourage more interest in this approach.
References in corpus (8)
- Auto-Encoding Variational Bayes
- Expectation Propagation for approximate Bayesian inference
- Gaussian Processes for Big Data
- Scalable Variational Gaussian Process Classification
- Variational Fourier features for Gaussian processes
- MCMC for Variationally Sparse Gaussian Processes
- Importance Weighting and Variational Inference
- Deep Gaussian Processes with Importance-Weighted Variational Inference
Cited by in corpus (18)
- Advanced Stationary and Non-Stationary Kernel Designs for Domain-Aware Gaussian Processes
- Kernel methods through the roof: handling billions of points efficiently
- Multioutput Gaussian Processes with Functional Data: A Study on Coastal Flood Hazard Assessment
- Nonnegative spatial factorization
- A Virtual Solar Wind Monitor at Mars with Uncertainty Quantification using Gaussian Processes
- GPflux: A Library for Deep Gaussian Processes
- Variational Orthogonal Features
- Accelerating Surface Composition Characterization of Thin-Film Materials Libraries using Multi-Output Gaussian Process Regression
- Bayesian Image Classification with Deep Convolutional Gaussian Processes
- Pathwise Conditioning of Gaussian Processes
- Empirical Models for Multidimensional Regression of Fission Systems
- Matérn Gaussian Processes on Graphs
- Scalable Gaussian Processes for Data-Driven Design using Big Data with Categorical Factors
- Discovering Diverse Athletic Jumping Strategies
- Deep Neural Networks as Point Estimates for Deep Gaussian Processes
- Lecture notes on Machine Learning applications for global fits
- Validating Gaussian Process Models with Simulation-Based Calibration
- Gaussian Processes on Hypergraphs