How does bayesopt fit a Gaussian process regression model to noisy data?

Asked by James Finley on 15 Jan 2019

Latest activity Commented on by Don Mathis on 17 Jan 2019

14 views (last 30 days)

Hi,

I am using bayesopt to optimise a non-deterministic objective function. I have set the ‘IsObjectiveDeterministic’ input argument to ‘false’, to reflect the stochastic nature of my objective function. My objective function features different levels of noise, depending on the input that is applied to the model.

My question is, does the Gaussian process regression model used in bayesopt assume a constant variance on the noise applied to objective function, or does the GPR model use a non-identically distributed noise for different data points in the observed data? If the latter case is true, how is the noise estimated for different inputs?

Many thanks

0 Comments

Products

MATLAB

Release

R2018b

2 Answers

Link

Answer by Don Mathis on 16 Jan 2019

Edited by Don Mathis on 16 Jan 2019

Accepted Answer

bayesopt uses fitrgp to fit the GP models, which assumes constant noise everywhere.

https://www.mathworks.com/help/stats/bayesian-optimization-algorithm.html

2 Comments

James Finley on 17 Jan 2019

Hi Don,

Many thanks for your response, you have have answered my question. I was also wondering how fitrgp estimates the variance for the noise in a non-determinisitc system?

Thank you

Don Mathis on 17 Jan 2019

That's part of the Gaussian Process learning algorithm, described here https://www.mathworks.com/help/stats/gaussian-process-regression-models.html

Link

Answer by Resul Al on 17 Jan 2019

Hi Don,

Is there a way to make fitrgp to estimate heteroscedastic noise, i.e noise variance is not constant everywhere?

Thank you.

1 Comment

Don Mathis on 17 Jan 2019

fitrgp provides no built-in way to do that. It may be possible to do it with a custom kernel function, but I'm not sure.

Products

MATLAB

Release

R2018b

You are now following this question

James Finley (view profile)