Table 2 Summary of thermal boundary conductance (TBC) prediction results on training and test set.

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

 

Training set

Test set

 

RMSE (GWm−2K−1)

MAPE

RMSE (GWm−2K−1)

MAPE

Crystalline–crystalline (426 GBs)

0.044

5.8%

0.056

6.2%

Crystalline–amorphous (802 GBs)

0.047

8.1%

0.073

12.5%

Total (1228 GBs)

0.046

7.3%

0.068

10.4%

  1. Root mean squared error (RMSE) and mean absolute percentage error (MAPE) were calculated as evaluation metrics.
Back to article page