Fig. 6: Prediction results of 75 crystalline–crystalline grain boundaries that are electroincally realistic. | npj Computational Materials

Fig. 6: Prediction results of 75 crystalline–crystalline grain boundaries that are electroincally realistic.

From: Robust combined modeling of crystalline and amorphous silicon grain boundary conductance by machine learning

Fig. 6

Panels a, c, and e demonstrate parity plots between calculated and ensemble-predicted values with error bars indicating standard deviations of the ensemble predictions, while b, d, and f show feature importance of thermal boundary conductance (TBC), electrical boundary conductance (EBC), and EBC/TBC respectively.

Back to article page