Fig. 4: Selection and evaluation of hardness prediction ML model by estimating the test error. | npj Computational Materials

Fig. 4: Selection and evaluation of hardness prediction ML model by estimating the test error.

From: Data-driven design of novel lightweight refractory high-entropy alloys with superb hardness and corrosion resistance

Fig. 4

a Average prediction MAE value of testing set for six different ML algorithms after 100 times modeling. b The average prediction error of each MLP model contains a feature subset. c Performance of the trained MLP model on both the training set and the testing sets. d SHAP analysis result of ML hardness model.

Back to article page