fig3

Figure 3. ML performance using FODs for predicting target attributes ΔE_TT and ΔE_ST. (A and B) Performance metrics (R² in orange, RMSE in blue) across eight ML algorithms; (C-F) Predicted vs. actual values for RF and XGBoost models on both training and testing samples, demonstrating the prediction accuracy for attributes ΔE_TT (C and D) and ΔE_ST (E and F). The diagonal dashed lines represent ideal 1:1 correspondence. ML: Machine learning; FODs: fragment orbital descriptors; R²: the coefficient of determination; RMSE: root mean square error; RF: random forest; XGBoost: eXtreme gradient boosting; LR: linear regression; KNN: K-nearest neighbors; DT: decision tree; SVR: support vector regression; MLP: multilayer perceptron.