Table 2 Statistical analysis for ABC and ABC-PSO FE techniques.

From: Generalizability of machine learning models for diabetes detection a study with nordic islet transplant and PIMA datasets

Statistical parameters

Microarray gene dataset

PIMA dataset

ABC

ABC-PSO

ABC

ABC-PSO

Diab-Pa

Non-Diab Pa

Diab Pa

Non-Diab Pa

Diab-Pa

Non-Diab Pa

Diab Pa

Non-Diab Pa

Mean

0.8224

0.1100

1.0536

0.6811

0.8295

0.1101

0.9536

0.1584

Variance

8.56 × 10− 5

0.0001

0.1262

0.1516

3.3 × 10− 6

2.8 × 10− 6

0.0068

0.0129

Skewness

0.8291

0.5982

0.3469

0.3782

− 0.1106

− 0.013

0.0095

1.0509

Kurtosis

141.406

121.6949

0.4598

0.3309

5.5313

8.0587

0.8741

2.1452

Pearson correlation coefficient

− 0.006

0.0006

0.0161

0.0039

− 0.022

0.0138

− 0.031

0.0177

Sample entropy

11.486

11.4866

11.4868

11.4868

5.4085

4.8492

8.0715

8.9687

Kolmogorov complexity36

11.486

11.4828

11.4862

11.4861

3.3306

2.9359

8.072

8.9682

Hijroth parameters–complexity mobility37

1.5029

0.0218

1.7137

0.1133

1.7093

0.2790

1.5901

0.6632

1.4819

0.8648

1.67

0.3165

1.6063

0.1163

1.4854

0.8699

Higuchi fractal dimension38

1.9993

2.0007

2.0006

2.0093

1.9811

1.9979

2.0081

1.9982

CCA

0.0721

0.1084

0.1256

0.1597