Fig. 5: Extendibility of the GBM-LASSO model for new oxide species. | npj Computational Materials

Fig. 5: Extendibility of the GBM-LASSO model for new oxide species.

From: Predicting densities and elastic moduli of SiO2-based glasses by machine learning

Fig. 5

a, b Reproduction of the non-linear effects of B2O3 on bulk and shear modulus in the a xB2O3-(100-x)SiO2 and (b) xB2O3-(30-x)Na2O-70SiO2 glasses in the training set. c Predictions from GBM-LASSO versus MD results on the test set. The test set is composed of 15 randomly selected compositions for the B2O3-containing multicomponent glasses (detailed information is listed in Supplementary Table 7) that are not included in the training dataset.

Back to article page