Kaggle机器学习初级的三种决策树

提示：文章写完后，目录可以自动生成，如何生成可参考右边的帮助文档

文章目录

• 前言
• 一、决策树
• 总结

---

前言

决策树

---

一、决策树

单棵决策树 DecisionTreeRegressor

# 决策树回归案例 - 波士顿房价预测
import pandas as pd
from sklearn.datasets import load_diabetes
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeRegressor, plot_tree
from sklearn.metrics import mean_squared_error, r2_score
import matplotlib.pyplot as plt

# 加载糖尿病数据集
diabetes = load_diabetes()
X = pd.DataFrame(diabetes.data, columns=diabetes.feature_names)
y = diabetes.target

# 数据预处理
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

# 创建并训练决策树模型
dt_reg = DecisionTreeRegressor(
    max_depth=3,  # 限制树深度防止过拟合
    min_samples_split=10,  # 节点最小样本数
    random_state=42
)
dt_reg.fit(X_train, y_train)

# 预测与评估
y_pred = dt_reg.predict(X_test)

print("决策树回归结果:")

随机森林RandomForestRegressor

# 随机森林回归案例 - 空气质量预测
import pandas as pd
from sklearn.datasets import fetch_openml
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_squared_error, r2_score
import matplotlib.pyplot as plt

# 加载空气质量数据集
data = fetch_openml(name='pm25', version=1, as_frame=True)
df = data.frame.dropna()
X = df.drop(columns=['target'])
y = df['target']

# 数据预处理
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

# 创建并训练随机森林模型
rf_reg = RandomForestRegressor(
    n_estimators=200,  # 树的数量
    max_depth=10,      # 单棵树的最大深度
    min_samples_leaf=5,  # 叶节点最小样本数
    random_state=42,
    n_jobs=-1  # 使用所有CPU核心
)
rf_reg.fit(X_train, y_train)

# 预测与评估
y_pred = rf_reg.predict(X_test)

print("\n随机森林回归结果:")

XGBRegressor（一般用于房价预测）

# XGBoost回归案例 - 自行车租赁预测
import pandas as pd
from sklearn.datasets import fetch_openml
from sklearn.model_selection import train_test_split
from xgboost import XGBRegressor
from sklearn.metrics import mean_squared_error, r2_score
import matplotlib.pyplot as plt
from sklearn.preprocessing import LabelEncoder

# 加载自行车租赁数据集
data = fetch_openml(name='boston', version=1, as_frame=True)
df = data.frame
X = df.drop(columns=['MEDV'])
y = df['MEDV']

# 数据预处理（处理分类变量）
for col in X.select_dtypes(include=['object']).columns:
    le = LabelEncoder()
    X[col] = le.fit_transform(X[col])

# 数据分割
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

# 创建并训练XGBoost模型
xgb_reg = XGBRegressor(
    n_estimators=500,       # 树的数量
    learning_rate=0.05,     # 学习率
    max_depth=6,            # 树的最大深度
    subsample=0.8,          # 样本抽样比例
    colsample_bytree=0.8,   # 特征抽样比例
    random_state=42,
    n_jobs=-1               # 使用所有CPU核心
)
xgb_reg.fit(X_train, y_train, 
            eval_set=[(X_test, y_test)], 
            early_stopping_rounds=20,  # 早停防止过拟合
            verbose=False)

# 预测与评估
y_pred = xgb_reg.predict(X_test)

print("\nXGBoost回归结果:")

---

总结

黑箱：一般是处于利益考量或者有上亿个参数人类不可学习等因素的模型，比如XGBoostRegressor。