fig4

Figure 4. The MSE of training and test set under various sampling methods by different ML models at training sizes of 0.2, 0.6 and 0.8, with the upper triangle legend representing test set MSE, and the lower triangle legend representing training set MSE. The MSE of the training and test sets for FPS with Set A and B displays more similar color patterns across various ML models, indicating that less overfitting is achieved in these models by implementing the FPS strategy in the designated feature spaces. MSE: Mean squared error; ML: machine learning; FPS: farthest point sampling.