Exercise M6.05

Module 6. Ensemble of models Hyperparameters tuning with ensemble methods
#1

The notebook says:

“First, create a histogram gradient boosting regressor. You can set the trees number to be large, and configure the model to use early-stopping.”

Two notes:

  • there could be a note here regarding needing to use from sklearn.experimental import enable_hist_gradient_boosting
  • when creating an HistGradientBoostingRegressor () you can’t actually set the number of estimators.

Am I missing something?

#2

Hi Lokijota,

When you create your HGBR you can fix a large number of tree with max_iter

1 Like
#3

We add the note in the first couple of notebooks and assess that one should start to be autonomous at this stage.

#4

Je me rappelle plus bien mais il était pas censé y avoir une erreur avec un message qui aide quand on oublie d’importer sklearn.experimental ? Ça semble pas le cas en tout cas :

In [1]: from sklearn.ensemble import HistGradientBoostingClassifier
---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-1-d6293110ba11> in <module>
----> 1 from sklearn.ensemble import HistGradientBoostingClassifier

ImportError: cannot import name 'HistGradientBoostingClassifier' from 'sklearn.ensemble' (/home/lesteve/miniconda3/lib/python3.9/site-packages/sklearn/ensemble/__init__.py)