Fig. 2: Results of 50 GA runs and the classification performance for phase prediction of HEAs. | npj Computational Materials

Fig. 2: Results of 50 GA runs and the classification performance for phase prediction of HEAs.

From: Elemental numerical descriptions to enhance classification and regression model performance for high-entropy alloys

Fig. 2

a, b The classification accuracy of the logistic regression (LR) model as a function of the number of iterations within 50 GA runs for classification I and classification II, respectively. The red solid line indicates the best performer. c, e is the margin of LR model to classify the SS and NSS HEAs based on the materials features defined from the numerical descriptions of elements and traditional empirical features, respectively. d, f is the margin of the LR model to classify FCC, BCC, and DP HEAs based on the features of the material defined from the numerical descriptions of elements and traditional empirical features, respectively. The larger symbols represent the 15 newly synthesized HEAs.

Back to article page