search

hyperopt / hyperopt-sklearn

Hyper-parameter optimization for sklearn
hyperopt.github.io/hyperopt-sklearn
Other
1.59k stars 272 forks source link

Example to use alternative Loss Function #125

Open dhbrand opened 5 years ago

dhbrand commented 5 years ago

Do you have an example of how to use a different loss function in the estimator? I want to use recall or f1 score from sklearn but they require the predicted y. Not sure how I would program that before the model has been chosen.

driesvr commented 5 years ago

There's an example in the pull request below: https://github.com/hyperopt/hyperopt-sklearn/pull/47