JuniorTheoryCommon

What does this program print?

Python

from sklearn.datasets import load_diabetes
from sklearn.ensemble import HistGradientBoostingRegressor

X, y = load_diabetes(return_X_y=True)
model = HistGradientBoostingRegressor(max_iter=300, random_state=0)
model.fit(X, y)
print(model.n_iter_)

300

Expecting fewer than 300 trees, as if early stopping were on: the default early_stopping="auto" turns it on only above 10,000 rows, and load_diabetes has 442.
Expecting 10 from n_iter_no_change=10: that is how many trees without improvement stop training, while n_iter_ counts the trees actually built.

What changes in the output if early_stopping=True is passed, and which rows does the model then hold out?
Why is the test set never used to decide when boosting stops?

Back to questions Open topic in roadmap →Answer it yourself — Quiz

Back to questions

JuniorTheoryCommon

What does this program print?

Python

from sklearn.datasets import load_diabetes
from sklearn.ensemble import HistGradientBoostingRegressor

X, y = load_diabetes(return_X_y=True)
model = HistGradientBoostingRegressor(max_iter=300, random_state=0)
model.fit(X, y)
print(model.n_iter_)

300

Expecting fewer than 300 trees, as if early stopping were on: the default early_stopping="auto" turns it on only above 10,000 rows, and load_diabetes has 442.
Expecting 10 from n_iter_no_change=10: that is how many trees without improvement stop training, while n_iter_ counts the trees actually built.

What changes in the output if early_stopping=True is passed, and which rows does the model then hold out?
Why is the test set never used to decide when boosting stops?

Back to questions Open topic in roadmap →Answer it yourself — Quiz