Repository navigation
Add TuRBO, trust-region Bayesian optimization (Eriksson et al. 2019) - #1584
Merged
Merged
Conversation
TuRBO-1 with a Matérn-5/2 Gaussian process, Thompson sampling and hand-rolled linear algebra; no dependencies. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
|
Interesting, did not know this solver. Is it superior to the classic solvers for high dimensional problems? I personally did not have success stories with bayesian optimization algorithms in the past but I might not have tuned them well enough. |
Collaborator
Author
|
@dschmitz89 : I mainly wanted it as a comparison to MCS-I'm with you on this is appears MCS is better. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
TuRBO-1 with a Matérn-5/2 Gaussian process, Thompson sampling and hand-rolled linear algebra; no dependencies.
Best f − f* so far against function calls, up to 500 calls per dimension (MCS one deterministic run, the others the median of 25 seeds). TuRBO makes the fastest early progress among the stochastic optimizers, then stalls near 10⁻³:
All 20 problems, for H5Web: turbo.h5.zip
🤖 Generated with Claude Code