Skip to content

Commit

Permalink
docs: update clusterfold-api reference
Browse files Browse the repository at this point in the history
Signed-off-by: Emmanuel Ferdman <[email protected]>
  • Loading branch information
emmanuel-ferdman committed Nov 12, 2024
1 parent 69e8740 commit 422e673
Showing 1 changed file with 1 addition and 1 deletion.
2 changes: 1 addition & 1 deletion docs/user-guide/cross-validation.md
Original file line number Diff line number Diff line change
Expand Up @@ -129,7 +129,7 @@ To use `GroupTimeSeriesSplit` with sklearn's [GridSearchCV](https://scikit-learn

## Cluster-Kfold

The [ClusterFoldValidation](clusterfold-api) object is a cross-validator that splits the data into `n_splits` folds, where each fold is determined by a clustering algorithm. This is not a common pattern, probably more like an anti-pattern really, but it might be useful when you want to make sure that the train and test sets are very distinct. This can be seen as a way to make it harder for the algorithm perform well, because the training sets are sampled differently than the test sets.
The [ClusterFoldValidation][clusterfold-api] object is a cross-validator that splits the data into `n_splits` folds, where each fold is determined by a clustering algorithm. This is not a common pattern, probably more like an anti-pattern really, but it might be useful when you want to make sure that the train and test sets are very distinct. This can be seen as a way to make it harder for the algorithm perform well, because the training sets are sampled differently than the test sets.

### Example

Expand Down

0 comments on commit 422e673

Please sign in to comment.