Skip to main content

Table 3 Impacts of the hierarchical-information based (HIB) technique on mLASSO and mEN based on leave-one-out cross-validation (LOOCV) on the human dataset

From: Sparse regressions for predicting and interpreting subcellular localization of multi-label proteins

Measures

mLASSO

mEN

 

without HIB

with HIB

without HIB

with HIB

OAA

0.729

0.748

0.743

0.742

OLA

0.820

0.846

0.836

0.825

Accuracy

0.814

0.833

0.827

0.821

Precision

0.859

0.874

0.869

0.866

Recall

0.857

0.879

0.870

0.860

F1

0.843

0.862

0.855

0.849

Micro F1

0.826

0.844

0.837

0.831

Macro F1

0.638

0.676

0.741

0.667

HL

0.029

0.027

0.028

0.029

  1. Note that OAA is the most stringent and objective among all the metrics. Data in bold represent the best result of the corresponding measures among all predictors