Table 1 Performance of the constructed models measured as root-mean-square error (RMSE, nm) for three data sets. QSPR: quantitative structure–property relationship, SVR: support vector regression.
From: Prediction of coating thickness for polyelectrolyte multilayers via machine learning
RMSE (nm) | QSPR Bagging regressor | QSPR SVR | Bagging regressor |
|---|---|---|---|
Training set | 46.7 | 76 | 50.9 |
Test set | 68.7 | 113.8 | 73.8 |
Validation set | 74.7 | 123.1 | 226.6 |
