Scoring

On top of scikit-learn scoring parameter options, julearn extends the functionality with more internal scorers and the possibility to define custom scorers.

Internal Scorers

Name (str)

Description

r2_corr

Pearson product-moment correlation coefficient (squared), as computed by numpy.corrcoef

Custom Scorers

julearn allows the user to define any function and use it as a scorer in the same way scikit-learn or julearn internal scorers work.

In the example Custom Scoring Function for Regression you can see how to make use of this functionality.