机器学习实战：端到端房价预测模型

pip install scikit-learn pandas matplotlib xgboost

from sklearn.datasets import fetch_california_housing
from sklearn.model_selection import train_test_split
import pandas as pd
import numpy as np

housing = fetch_california_housing()
X = pd.DataFrame(housing.data, columns=housing.feature_names)
y = housing.target  # 房价中位数（单位：十万美元）

print(X.shape, y.shape)
print(X.head())

print(X.isnull().sum())  # 无缺失
print(X.describe())

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42
)

import matplotlib.pyplot as plt
plt.hist(y, bins=50)
plt.title("房价分布")
plt.show()

# 构造新特征：每间房的平均房间数、人口密度等
X_train['rooms_per_household'] = X_train['AveRooms'] / X_train['AveOccup']
X_train['bedrooms_per_room'] = X_train['AveBedrms'] / X_train['AveRooms']
X_train['population_per_household'] = X_train['Population'] / X_train['AveOccup']

X_test['rooms_per_household'] = X_test['AveRooms'] / X_test['AveOccup']
X_test['bedrooms_per_room'] = X_test['AveBedrms'] / X_test['AveRooms']
X_test['population_per_household'] = X_test['Population'] / X_test['AveOccup']

from sklearn.linear_model import LinearRegression
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import cross_val_score

models = {
    "线性回归": LinearRegression(),
    "决策树": DecisionTreeRegressor(random_state=42),
    "随机森林": RandomForestRegressor(n_estimators=100, random_state=42)
}

for name, model in models.items():
    scores = cross_val_score(model, X_train, y_train, cv=5, 
                             scoring='neg_mean_squared_error')
    rmse = np.sqrt(-scores.mean())
    print(f"{name}: RMSE = {rmse:.4f}")

from xgboost import XGBRegressor
xgb = XGBRegressor(n_estimators=100, learning_rate=0.1, random_state=42)
scores = cross_val_score(xgb, X_train, y_train, cv=5, scoring='neg_mean_squared_error')
print(f"XGBoost: RMSE = {np.sqrt(-scores.mean()):.4f}")

from sklearn.model_selection import GridSearchCV

param_grid = {
    'n_estimators': [100, 200],
    'max_depth': [None, 10, 20],
    'min_samples_split': [2, 5]
}

rf = RandomForestRegressor(random_state=42)
grid_search = GridSearchCV(rf, param_grid, cv=3, 
                          scoring='neg_mean_squared_error', n_jobs=-1)
grid_search.fit(X_train, y_train)

print("最佳参数:", grid_search.best_params_)
best_model = grid_search.best_estimator_

from sklearn.metrics import mean_squared_error, r2_score

y_pred = best_model.predict(X_test)
rmse = np.sqrt(mean_squared_error(y_test, y_pred))
r2 = r2_score(y_test, y_pred)
print(f"测试集 RMSE: {rmse:.4f}, R²: {r2:.4f}")

feature_importances = pd.Series(
    best_model.feature_importances_, index=X_train.columns
).sort_values(ascending=False)

print(feature_importances.head(10))
plt.figure(figsize=(8,5))
feature_importances.head(10).plot(kind='bar')
plt.title('特征重要性 Top 10')
plt.show()