======================================================================
模型训练报告
======================================================================

1. 数据概览:
   - 总样本数: 2941
   - 训练集: 2058 (70%)
   - 测试集: 883 (30%)
   - 特征数: 16

2. 模型评估结果:
               模型          MSE         RMSE          MAE       R²    MAPE(%)  训练时间(秒)
Linear Regression 5.696363e+08 23867.055529 15987.923799 0.795971 487.795101 0.000497
 Ridge Regression 5.699472e+08 23873.566674 15987.987540 0.795860 487.437312 0.000322
    Random Forest 2.695686e+06  1641.854346   893.146256 0.999034   5.505826 0.174586
Gradient Boosting 2.842279e+06  1685.906038   958.002559 0.998982   7.354316 0.427650

3. 交叉验证结果:
               模型     R²均值    R²标准差        MSE均值       MSE标准差
Linear Regression 0.797486 0.011782 5.210984e+08 4.479025e+07
 Ridge Regression 0.797500 0.011821 5.210798e+08 4.502908e+07
    Random Forest 0.998227 0.000199 4.553287e+06 5.216479e+05
Gradient Boosting 0.998740 0.000253 3.283128e+06 8.355211e+05

4. Top 10 重要特征:
                      Linear Regression  Ridge Regression  Random Forest  Gradient Boosting
unit_price                     0.563122          0.562381       0.667421           0.665657
quantity                       0.299895          0.300142       0.325300           0.331876
cost                           0.006870          0.008225       0.005761           0.002358
price_cost_ratio               0.009286          0.008893       0.000218           0.000025
expected_profit_rate           0.001943          0.002229       0.000217           0.000022
order_month                    0.026363          0.025930       0.000204           0.000013
order_dayofweek                0.017439          0.017422       0.000176           0.000008
region_avg_price               0.011758          0.011787       0.000167           0.000009
region_encoded                 0.008188          0.008200       0.000135           0.000006
order_quarter                  0.030632          0.030229       0.000095           0.000002

======================================================================
训练完成！
======================================================================