Subspace-Induced Gaussian Processes

02/21/2018
by   Zilong Tan, et al.
0

We present a new Gaussian process (GP) regression model where the covariance kernel is indexed or parameterized by a sufficient dimension reduction subspace of a reproducing kernel Hilbert space. The covariance kernel will be low-rank while capturing the statistical dependency of the response to the covariates, this affords significant improvement in computational efficiency as well as potential reduction in the variance of predictions. We develop a fast Expectation-Maximization algorithm for estimating the parameters of the subspace-induced Gaussian process (SIGP). Extensive results on real data show that SIGP can outperform the standard full GP even with a low rank-m, m≤ 3, inducing subspace.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/13/2021

How Good are Low-Rank Approximations in Gaussian Process Regression?

We provide guarantees for approximate Gaussian Process (GP) regression r...
research
07/26/2022

Large-Scale Low-Rank Gaussian Process Prediction with Support Points

Low-rank approximation is a popular strategy to tackle the "big n proble...
research
10/31/2017

Tensor Regression Meets Gaussian Processes

Low-rank tensor regression, a new model class that learns high-order cor...
research
02/27/2017

Embarrassingly parallel inference for Gaussian processes

Training Gaussian process (GP)-based models typically involves an O(N^3...
research
12/17/2021

GP-HMAT: Scalable, O(nlog(n)) Gaussian Process Regression with Hierarchical Low-Rank Matrices

A Gaussian process (GP) is a powerful and widely used regression techniq...
research
05/23/2016

Collaborative Filtering with Side Information: a Gaussian Process Perspective

We tackle the problem of collaborative filtering (CF) with side informat...
research
05/08/2018

Spatial shrinkage via the product independent Gaussian process prior

We study the problem of sparse signal detection on a spatial domain. We ...

Please sign up or login with your details

Forgot password? Click here to reset