Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Surrogate residuals from an arbitrary sample #10

Open
cbujard opened this issue Mar 19, 2024 · 0 comments
Open

Surrogate residuals from an arbitrary sample #10

cbujard opened this issue Mar 19, 2024 · 0 comments

Comments

@cbujard
Copy link

cbujard commented Mar 19, 2024

We should be able to get the residuals from the current surrogate evaluated on an aribtrary (test) sample (for which we know the true values). The following instruction evaluates a "metric" based on the provided sample file (which contains both the input parameters and their corresponding true values). Any of these metrics are in fact directly computed from the residuals, but there seem to be no way to actually obtain these residuals. In the special case of gaussian process surrogates, it would also be desirable to not only get the residual at a given input point, but also the surrogate uncertainty at this point.

https://snl-dakota.github.io/docs/6.18.0/users/usingdakota/reference/model-surrogate-global-import_challenge_points_file.html

While these metrics offer a basic mean to evaluate the accuracy of a surrogate, these are not generally sufficient and more work based on such testing residuals is often needed.

@cbujard cbujard changed the title Surrogate residuals from an arbitrary (user defined) sample Surrogate residuals from an arbitrary sample Mar 19, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant