Table 1.
Comparison of regression models using 5-fold cross-validation.
Method | RMSE |
---|---|
Linear regression | 0.23 |
Robust linear regression | 0.24 |
Fine tree | 0.05 |
Medium tree | 0.06 |
Coarse tree | 0.08 |
SVM with linear kernel | 0.24 |
SVM with quadratic kernel | 0.14 |
SVM with cubic kernel | 0.7 |
SVM with fine Gaussian kernel | 0.14 |
SVM with medium Gaussian kernel | 0.8 |
SVM with coarse Gaussian kernel | 0.21 |
Boosted tree based ensemble | 0.16 |
Bagged tree based ensemble | 0.03 |