Skip to main content
. 2020 Jun 15;8(3):215–250. doi: 10.5599/admet.835

Table 4.

Statistical evaluations of predictions for the Extra Trees, Random Forest and Bagging predictors and for the Vox Machinarum consensus classifier over Avdeef’s average logS values [45] for all 100 compounds comprising the 2019 Solubility Challenge tight test set of 100 molecules. The Vox Machinarum predictions reported here were the median of the other three classifiers’ predictions for each compound. The standard deviation of the 100 compounds’ logS values was 1.266.

Method RMSE RMSE/SD AAE R 2 Err < 0.5 Err < 1.0
Extra Trees 0.946 0.748 0.720 0.527 45 (45%) 75 (75%)
Random Forest 0.989 0.781 0.765 0.494 44 (44%) 70 (70%)
Bagging 1.023 0.808 0.815 0.481 38 (38%) 65 (65%)
Vox Machinarum 0.977 0.771 0.754 0.507 46 (46%) 69 (69%)