Table 5.
Training Set (%) | Score | r 2 | SD | q 2 | RMSE | N | Fingerprint |
---|---|---|---|---|---|---|---|
70 | 0.85 | 0.85 | 0.40 | 0.82 | 0.37 | 2 | Desc |
72 | 0.90 | 0.89 | 0.32 | 0.90 | 0.29 | 2 | Dendritic |
74 | 0.82 | 0.82 | 0.44 | 0.78 | 0.39 | 2 | Desc |
76 | 0.81 | 0.81 | 0.43 | 0.78 | 0.40 | 2 | Desc |
78 | 0.81 | 0.81 | 0.43 | 0.78 | 0.40 | 2 | Desc |
80 | 0.82 | 0.82 | 0.43 | 0.91 | 0.26 | 2 | Desc |
Score: Performance considering training and test set predictive accuracies; r2: Coefficient of determination for the training set; SD: Standard deviation; q2: Predictive correlation coefficient for the test set (r²pred); RMSE: Root mean square error for the test set predictions; N: Optimum number of components.