有监督式决策树
from sklearn.ensemble import RandomForestClassifier
from sklearn.datasets import make_classification
X, y = make_classification(n_samples=1000, n_features=4,
n_informative=2, n_redundant=0,
random_state=0, shuffle=False)
clf = RandomForestClassifier(n_estimators=100, max_depth=2,
random_state=0)
clf.fit(X, y)
print(clf.feature_importances_)
print(clf.predict([[0, 0, 0, 0]]))