Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

Comments¶

1. If you want a simple decision tree,

DecisionTreeRegressor is prefer to DecisionTreeClassifier as DecisionTreeRegressor is more flexible.

As a matter of fact, ensemble tree-based models (no matter classifiers or regressors) are based on DecisionTreeRegressor instead of DecisionTreeClassifier .

Visualize Decision Trees¶

What is the meaning of "value" in a node in sklearn decisiontree plot_tree

1. DecisionTreeClassifier: number of samples in each node

   DecisionTreeRegressor: predicted value of the node, i.e., the $v_m$ parameter ...

References¶

• staged_predict

• staged_predict_prob

• GradientBoostingClassifier