fig4

Figure 4. Comparative performance of FODs vs. MODs across eight ML models. Radar plots display R² values for predictions of (A) ΔE_TT and (B) ΔE_ST, where radial distance from the center corresponds to prediction accuracy (greater R² values extend further toward the outer edge). Each axis represents one ML algorithm, enabling direct visual comparison of descriptor effectiveness. FODs: Fragment orbital descriptors; MODs: molecule orbital descriptors of complete DAD molecules; ML: machine learning; R²: the coefficient of determination; RF: random forest; MLP: multilayer perceptron; SVR: support vector regression; DT: decision tree; KNN: K-nearest neighbors; LR: linear regression; XGBoost: eXtreme gradient boosting.