Efficient leave-one-out cross-validation for Gaussian processes #3098
Conversation
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
meta-cla
bot
added
the
CLA Signed
|
@SebastianAment has exported this pull request. If you are a Meta employee, you can view the originating Diff in D88273413. |
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 3, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
force-pushed
the
export-D88273413
branch
from
5aa2693 to
ee39085
Compare
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #3098 +/- ##
=======================================
Coverage 99.99% 99.99%
=======================================
Files 219 219
Lines 21013 21080 +67
=======================================
+ Hits 21012 21079 +67
Misses 1 1 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 3, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
force-pushed
the
export-D88273413
branch
2 times, most recently
from
f289b11 to
3b54e7b
Compare
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 8, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 8, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 8, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 8, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
SebastianAment
force-pushed
the
export-D88273413
branch
from
3b54e7b to
9e6b37e
Compare
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 8, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Differential Revision: D88273413
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Reviewed By: Balandat Differential Revision: D88273413
SebastianAment
force-pushed
the
export-D88273413
branch
from
9e6b37e to
9b23352
Compare
SebastianAment
added a commit
to SebastianAment/botorch
that referenced
this pull request
Dec 9, 2025
…-pytorch#3098) Summary: BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately. However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total. This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals. Reviewed By: Balandat Differential Revision: D88273413
|
This pull request has been merged in 976a944. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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.
Summary:
BoTorch currently supports a naive version of leave-one-out (LOO) cross-validation, which entails generating all folds directly and evaluating the model on all folds separately.
However, for Gaussian processes, there is an efficient way to compute all LOO results (LOO predictive means and variances) with a single linear-algebraic formula with a O(n^3) complexity, whereas the naive approach takes O(n^3) per fold, so O(n^4) in total.
This commit implements the effient LOO CV method for non-ensemble BoTorch models whose prior marginal distributions are a multi-variate Normals.
Differential Revision: D88273413