Question

如何在基于决策树的分类方法中使用修剪和提升？

I have 10 features and 3000 samples.

Answer 1

以下是演示如何使用Boosting的示例。

from sklearn.datasets import make_classification
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.tree import DecisionTreeClassifier
from sklearn.cross_validation import StratifiedShuffleSplit
from sklearn.metrics import classification_report

# generate some artificial data
X, y = make_classification(n_samples=3000, n_features=10, n_informative=2, flip_y=0.1, weights=[0.15, 0.85], random_state=0)

# train/test split
split = StratifiedShuffleSplit(y, n_iter=1, test_size=0.2, random_state=0)
train_index, test_index = list(split)[0]
X_train, y_train = X[train_index], y[train_index]
X_test, y_test = X[test_index], y[test_index]

# boosting: many many weak classifiers (max_depth=1) refine themselves sequentially
# tree is the default the base classifier
estimator = GradientBoostingClassifier(n_estimators=200, learning_rate=0.1, max_depth=1, random_state=0)
estimator.fit(X_train, y_train)
y_pred = estimator.predict(X_test)
print(classification_report(y_test, y_pred))

             precision    recall  f1-score   support

          0       0.88      0.80      0.84       109
          1       0.96      0.98      0.97       491

avg / total       0.94      0.94      0.94       600

# benchmark: a standard tree
tree_benchmark = DecisionTreeClassifier(max_depth=3, class_weight='auto')
tree_benchmark.fit(X_train, y_train)
y_pred_benchmark = tree_benchmark.predict(X_test)
print(classification_report(y_test, y_pred_benchmark))

             precision    recall  f1-score   support

          0       0.63      0.84      0.72       109
          1       0.96      0.89      0.92       491

avg / total       0.90      0.88      0.89       600

决策树中的修剪与提升

1 个答案: