Activity Number:
|
432
- Contributed Poster Presentations: Royal Statistical Society
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 10, 2022 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Royal Statistical Society
|
Abstract #323024
|
|
Title:
|
O(N) Approximate Bayesian Gaussian Process Regression
|
Author(s):
|
Kelly R Moran*
|
Companies:
|
Los Alamos National Laboratory
|
Keywords:
|
Big data;
Bayesian Gaussian process;
Kernel approximation;
Scalability;
Surface estimation
|
Abstract:
|
Gaussian processes (GPs) are common components in Bayesian non-parametric models having a rich methodological literature and strong theoretical grounding. The use of exact GPs in Bayesian models is limited to problems containing several thousand observations due to their prohibitive computational demands. We leverage our novel posterior sampling algorithm along with kernel approximation methods to develop O(n) fully Bayesian Gaussian process regression to problems having an input space of dimension up to 3. We show that this approximation's Kullback-Leibler divergence to the true posterior can be made arbitrarily small. This work represents an improvement on our previous scaling in inputs, in which d-dimensional surfaces were modeled as tensor products of univariate GPs. We illustrate the performance of this fast increased fidelity approximate GP using both simulated and real data sets.
|
Authors who are presenting talks have a * after their name.