回归模型评估:MSE, RMSE, MAE, MAPE, R²
Excel 文件里有两列数据,分别是某地区房屋价格的预测值和真实值,对其进行误差评估
SalePrice_predictvalue |
SalePrice |
205578.0852 |
208500 |
176323.4192 |
181500 |
217021.2719 |
223500 |
161173.8857 |
140000 |
281214.8513 |
250000 |
… |
… |
A |
|
1 |
=T("houseprice_result.xls") |
2 |
=mse(A1.(SalePrice_predictvalue),A1.(SalePrice)) |
3 |
=sqrt(A2) |
4 |
=mae(A1.(SalePrice_predictvalue),A1.(SalePrice)) |
5 |
=A1.(abs((SalePrice_predictvalue-SalePrice)*100/SalePrice)).avg() |
6 |
=1-A2/var(A1.(SalePrice)) |
A1将文件读为序表
A2计算MSE,均方误差
A3 计算RMSE,均方根误差
A4计算MAE,平均绝对误差
A5 计算MAPE,平均绝对百分比误差
A6 计算R²,拟合优度