Data

Model

Normalization

None

Z-standard

Min-Max

MaxAbs (−1, 1)

Quantile Transform

Quantile Normalize

Arrhythmia data with binary target

Logistic

Total Loss

0.065

0.875

0.755

0.636

0.905

0.064

Bias

0.046

0.828

0.650

0.476

0.877

0.046

Variance

0.018

0.047

0.105

0.144

0.027

0.018

Noise

0.000

0.000

0.000

0.016

0.000

0.000

Variance-Bias Ratio

0.398

0.057

0.161

0.302

0.031

0.395

Percent Change from Raw

~

1350.982

1165.815

983.032

1397.399

98.683

Decision Tree

Total Loss

0.034

0.094

0.099

0.098

0.100

0.034

Bias

0.018

0.056

0.055

0.056

0.055

0.018

Variance

0.016

0.038

0.044

0.046

0.044

0.016

Noise

0.000

0.000

0.000

0.004

0.000

0.000

Variance-Bias Ratio

0.907

0.691

0.796

0.834

0.798

0.907

Percent Change from Raw

~

274.608

290.841

286.111

291.056

100.000

Random Forest

Total Loss

0.059

0.111

0.172

0.153

0.160

0.059

Bias

0.059

0.057

0.063

0.062

0.062

0.059

Variance

0.001

0.054

0.108

0.127

0.098

0.001

Noise

0.000

0.000

0.000

0.035

0.000

0.000

Variance-Bias Ratio

0.013

0.937

1.708

2.057

1.594

0.013

Percent Change from Raw

~

186.609

289.307

258.218

268.824

100.000

SVM

Total Loss

0.059

0.059

0.059

0.059

0.059

0.059

Bias

0.059

0.059

0.059

0.059

0.059

0.059

Variance

0.000

0.000

0.000

0.000

0.000

0.000

Noise

0.000

0.000

0.000

0.000

0.000

0.000

Variance-Bias Ratio

0.000

0.000

0.000

0.000

0.000

0.000

Percent Change from Raw

~

100.000

100.000

100.000

100.000

100.000