決定木のパラメータ | 複雑さを制御して汎化させる

import numpy as np
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeRegressor
from sklearn.datasets import make_regression
from sklearn.metrics import r2_score

X, y = make_regression(
    n_samples=500,
    n_features=2,
    noise=0.2,
    random_state=42,
)
Xtr, Xte, ytr, yte = train_test_split(X, y, test_size=0.3, random_state=0)

def evaluate(params):
    model = DecisionTreeRegressor(random_state=0, **params).fit(Xtr, ytr)
    r2_train = r2_score(ytr, model.predict(Xtr))
    r2_test = r2_score(yte, model.predict(Xte))
    print(f"{params}: train R2={r2_train:.3f}, test R2={r2_test:.3f}")

evaluate({"max_depth": 3})
evaluate({"max_depth": 10})
evaluate({"max_depth": 5, "min_samples_leaf": 5})
evaluate({"max_depth": 5, "ccp_alpha": 0.01})