Skip to main content

Table 1 AUROC comparison among seven algorithms based on multi-encodings by independent test

From: NmSEER V2.0: a prediction tool for 2′-O-methylation sites based on random forest and multi-encoding combination

Algorithm\Encoding

One-hot

PSNSP

PSDSP

KNF

KSNPF

RF

0.811

0.804

0.850

0.761

0.722

LR

0.761

0.762

0.833

0.761

0.740

GaussianNB

0.756

0.753

0.824

0.729

0.646

Adaboost

0.756

0.765

0.847

0.747

0.704

SVM

0.810

0.747

0.831

0.686

0.730

MLP

0.810

0.777

0.844

0.758

0.759

CNN

0.822

0.771

0.842

0.749

0.653

RF (1:1 training set)

0.801

0.797

0.848

0.763

0.724