19-Mar-2024, 11:02:31
---------------------------------------------------------
Traditional ML algorithm Performance
Dataset Size = 1000
c term coefficient = 0.024316352791230268
bv term coefficient = 4.579272409468922
ang term coefficient = 0.09347620772243007
***************
Training set
Set size = 600
R2 = 0.923
MAE = 0.130
RMSE = 0.165
Mean Bias Error = 0.000
Mean Absolute Percentage Error = 0.045
Error Standard Deviation = 0.165
***************
Validation set
Set size = 200
R2 = 0.936
MAE = 0.125
RMSE = 0.159
Mean Bias Error = -0.003
Mean Absolute Percentage Error = 0.044
Error Standard Deviation = 0.159
***************
Test set
Set size = 200
R2 = 0.930
MAE = 0.133
RMSE = 0.178
Mean Bias Error = -0.004
Mean Absolute Percentage Error = 0.047
Error Standard Deviation = 0.178
---------------------------------------------------------
OUTLIERS (TEST SET)
01) 23.vasp    Error: -0.564 eV    (index=44)
02) 46.vasp    Error: -0.673 eV    (index=86)
03) 625.vasp    Error: -0.794 eV    (index=123)
