Journal of Cheminformatics

Table 3 Prediction of the DFT DM by random forests using DM_NBO or DM_PEOE

From: Machine learning for the prediction of molecular dipole moments obtained by density functional theory

Descriptors	Training set^a		Test set
Descriptors	MAE (D)	R²/RMSE (D)	MAE (D)	R²/RMSE (D)
RDF_N + DM ^b_NBO	0.639	0.747/0.930	0.638	0.761/0.929
RDF_P + DM ^c_PEOE	0.624	0.740/0.946	0.615	0.765/0.929
PchmDM_N + DM ^b_NBO	0.647	0.753/0.924	0.651	0.769/0.921
PchmDM_P + DM ^c_PEOE	0.639	0.735/0.953	0.630	0.759/0.936
CDK + DM ^d_NBO	0.713	0.705/1.00	0.700	0.724/0.990
CDK + DM ^e_PEOE	0.708	0.685/1.03	0.704	0.705/1.02
MACCS + DM_NBO	0.526	0.806/0.813	0.507	0.826/0.792
MACCS + DM_PEOE	0.563	0.777/0.873	0.543	0.801/0.847

^aOOB estimation
^bDescriptors calculated using NBO charges, and DM_NBO
^cDescriptors calculated using PEOE charges, and DM_PEOE
^dGeometric CDK, and DM_NBO
^eGeometric CDK, and DM_PEOE

Back to article page

ISSN: 1758-2946

Contact us

Submission enquiries: journalsubmissions@springernature.com