机器学习方法

from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split
from sklearn.metrics import r2_score, mean_absolute_error

# 特征工程
df['hour'] = df.index.hour
df['day_of_week'] = df.index.dayofweek
df['is_weekend'] = df['day_of_week'].isin([5, 6]).astype(int)
df['is_peak_hour'] = df['hour'].isin([7, 8, 9, 17, 18, 19]).astype(int)

# 准备特征和目标变量
features = ['hour', 'day_of_week', 'is_weekend', 'is_peak_hour',
            'speed', 'occupancy', 'volume_lag_1', 'volume_lag_2']
X = df[features]
y = df['volume']

# 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
)

# 训练随机森林模型
rf_model = RandomForestRegressor(
    n_estimators=100,
    max_depth=10,
    min_samples_split=5,
    random_state=42
)
rf_model.fit(X_train, y_train)

# 预测和评估
y_pred = rf_model.predict(X_test)
r2 = r2_score(y_test, y_pred)
mae = mean_absolute_error(y_test, y_pred)

print(f"R² 分数: {r2:.4f}")
print(f"MAE: {mae:.2f}")

# 特征重要性分析
importance = pd.DataFrame({
    'feature': features,
    'importance': rf_model.feature_importances_
}).sort_values('importance', ascending=False)
print("\n特征重要性：")
print(importance)

import xgboost as xgb
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import mean_squared_error

# 创建XGBoost数据对象
dtrain = xgb.DMatrix(X_train, label=y_train)
dtest = xgb.DMatrix(X_test, label=y_test)

# 参数网格搜索
param_grid = {
    'max_depth': [4, 6, 8],
    'learning_rate': [0.01, 0.05, 0.1],
    'n_estimators': [100, 200, 300],
    'subsample': [0.8, 0.9, 1.0]
}

# 网格搜索
xgb_model = xgb.XGBRegressor(
    objective='reg:squarederror',
    random_state=42
)
grid_search = GridSearchCV(
    xgb_model, param_grid, cv=3,
    scoring='neg_mean_squared_error',
    n_jobs=-1
)
grid_search.fit(X_train, y_train)

# 最佳模型
best_model = grid_search.best_estimator_
print(f"最佳参数: {grid_search.best_params_}")

# 预测
y_pred = best_model.predict(X_test)
rmse = np.sqrt(mean_squared_error(y_test, y_pred))
print(f"RMSE: {rmse:.2f}")

# 预测结果
# 最佳参数: learning_rate=0.05, max_depth=6, n_estimators=200, subsample=0.9
# RMSE: 45.23