Table 1 Error statistics for forward prediction.

From: Attribute driven inverse materials design using deep learning Bayesian framework

 

HOMO

LUMO

Oxidation

Reduction

 

Train

Test

Train

Test

Train

Test

Train

Test

\({R}^{2}\)

0.9184

0.9164

0.9506

0.9497

0.9109

0.9070

0.9203

0.9160

MAE (eV/V)

0.1407

0.1435

0.1464

0.1476

0.1496

0.1510

0.1782

0.1808

\(<0.05\) (eV/V)

27.27

20.78

26.4

21.07

25.78

26.05

23.00

22.80

\(>0.5\) (eV/V)

2.07

3.62

2.64

4.64

2.96

3.05

5.71

6.07

  1. This table lists coefficient of determination (\({R}^{2}\)) and mean absolute error (MAE) of deep learning predictions. Percentage of molecules with absolute error \(<0.05\) eV/V and \(>0.5\) eV/V are also reported in the table
Back to article page