Table 3 Adding chemical shift information to concentration information provides the best classification of samples in the MTBLS374 dataset.

From: Improving sample classification by harnessing the potential of 1H-NMR signal chemical shifts

 

Both sets of information

Concentration information

Chemical shift information

Accuracy

0.899

0.806

0.883

kappa

0.797

0.614

0.766

AUROC

0.950

0.856

0.937

  1. Several quality indicators of the models generated only with concentration information, only with chemical shift information and with both sources of information are shown.
Back to article page