JuniorTheoryCommon

What does this program print?

Python

from sklearn.tree import DecisionTreeClassifier

X = [[1], [2], [3], [4], [5], [6]]
y = [0, 0, 1, 0, 1, 1]
deep = DecisionTreeClassifier(random_state=0).fit(X, y)
stump = DecisionTreeClassifier(max_depth=1, random_state=0).fit(X, y)
print(deep.score(X, y), round(stump.score(X, y), 2))

1.0 0.83

Expecting the unlimited tree to miss the lone 0 at x = 4: without max_depth it keeps splitting until every leaf is pure, so it scores 1.0 on its training data.
Expecting max_depth=1 to make no split: depth 1 is one question and two leaves, which gets 5 of the 6 training samples right, so the score is 0.83.

Which of the two trees would you expect to do better on new data, and why?
How does min_samples_leaf limit a tree differently from max_depth?

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

Back to questions

JuniorTheoryCommon

What does this program print?

Python

from sklearn.tree import DecisionTreeClassifier

X = [[1], [2], [3], [4], [5], [6]]
y = [0, 0, 1, 0, 1, 1]
deep = DecisionTreeClassifier(random_state=0).fit(X, y)
stump = DecisionTreeClassifier(max_depth=1, random_state=0).fit(X, y)
print(deep.score(X, y), round(stump.score(X, y), 2))

1.0 0.83

Expecting the unlimited tree to miss the lone 0 at x = 4: without max_depth it keeps splitting until every leaf is pure, so it scores 1.0 on its training data.
Expecting max_depth=1 to make no split: depth 1 is one question and two leaves, which gets 5 of the 6 training samples right, so the score is 0.83.

Which of the two trees would you expect to do better on new data, and why?
How does min_samples_leaf limit a tree differently from max_depth?

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