Skip to main content

Table 2 Accuracy measures for simulated data. Accuracy measures obtained with different classification rules for simulated data with 10 variables (simulated independently from a Gaussian distribution with zero mean and unit variance), 1,000 samples and 2 CV folds

From: Joint use of over- and under-sampling techniques and cross-validation for the development and assessment of prediction models

 

PA

PAmin

PAmaj

GM

F1

AUC

Under (incorrect)

0.5003

0.5015

0.4991

0.4978

0.4996

0.5005

Under (correct)

0.4999

0.5009

0.4998

0.4985

0.1668

0.5003

Over (incorrect)

0.7624

0.8438

0.6809

0.7575

0.7801

0.7986

Over (correct)

0.7191

0.2254

0.7740

0.4151

0.1380

0.5004

100-SM (incorrect)

0.6422

0.7492

0.4281

0.5644

0.7359

0.5858

200-SM (incorrect)

0.6565

0.7462

0.5220

0.6230

0.7224

0.6431

500-SM (incorrect)

0.7012

0.7715

0.6168

0.6892

0.7378

0.7148

100-SM (correct)

0.4520

0.5582

0.4402

0.4940

0.1692

0.4993

200-SM (correct)

0.5294

0.4629

0.5368

0.4966

0.1643

0.4998

500-SM (correct)

0.5997

0.3757

0.6245

0.4824

0.1579

0.5007