Table 2.
Prediction accuracy and cardinality for the best ten models obtained by Soto’s method [5]
| Model | Predictive accuracy | Cardinality |
|---|---|---|
| M1 (Mn/MW, Sp, RHyDp, ETA_EtaP_F_L) | R2 = 0.26 MAE = 4.62 RMSE = 8.14 |
4 |
| M2 (Mn/MW, MDEO-11, D/Dr09, SMTIV) | R2 = 0.32 MAE = 5.94 RMSE = 8.31 |
4 |
| M3 (Mn/MW, nHBint4, nHBint10, ETA_dEpsilon_B) | R2 = 0.56 MAE = 4.03 RMSE = 6.22 |
4 |
| M4 (Mn/MW, nsCH3, nF6Ring, ALOGP2, RDCHI) | R2 = 0.41 MAE = 3.94 RMSE = 6.75 |
5 |
| M5 (Mn/MW, nROH, n6Ring, nHCsatu, ALOGP2) | R2 = 0.68 MAE = 3.28 RMSE = 5.78 |
5 |
| M6 (Mn/MW,nP, minHBa, T(O..P), ETA_Epsilon_3) | R2 = 0.25 MAE = 4.48 RMSE = 7.20 |
5 |
| M7 (Mn/MW, ETA_dEpsilon_B, C-005, SHaaCH, nHBint9,nCt) | R2 = 0.31 MAE = 4.19 RMSE = 7.20 |
6 |
| M8 (Mn/MW, ndssC, minHBint9, MSD, C-004, Mw/Mn (PDI), crosshead speed(CHS)) | R2 = 0.39 MAE = 3.92 RMSE = 6.86 |
7 |
| M9 (Mn/MW, Pol, Wap, maxHAvin, nHAvin, MWC04) | R2 = 0.15 MAE = 4.92 RMSE = 7.88 |
6 |
| M10 (Mn/MW,maxHBint6, ETA_dEpsilon_A, TIC2, ndO, nHdCH2) | R2 = 0.48 MAE = 4.02 RMSE = 7.09 |
6 |
The second column shows the predictive accuracy of the “best” model after applying 4-fold cross validation on three different methods (linear regression, decision trees, and neural networks). The parameter setup and predictive accuracy for all methods is available in the Additional file 1: Table S2.