Skip to main content

Table 6 Dunn’s post-hoc test between models

From: Application of deep learning and feature selection technique on external root resorption identification on CBCT images

Sample 1- Sample 2

Test Statistic

Std. Error

Std. Test Statistic

Sig.

Adj. Sig. a

RF + EFNET-FS + RF + EFNET

-2.500

8.065

-0.310

0.757

1.000

RF + EFNET-SVM + EFNET

-7.250

8.065

-0.899

0.369

1.000

RF + EFNET-FS + SVM + EFNET

-12.583

8.065

-1.560

0.119

1.000

RF + EFNET-SVM + VGG

-16.000

8.065

-1.984

0.047

1.000

RF + EFNET-RF + VGG

19.417

8.065

2.408

0.016

0.450

RF + EFNET-FS + SVM + VGG

-21.667

8.065

-2.687

0.007

0.202

RF + EFNET-FS + RF + VGG

-26.583

8.065

-3.296

0.001

0.027*

FS + RF + EFNET-SVM + EFNET

4.750

8.065

0.589

0.556

1.000

FS + RF + EFNET-FS + SVM + EFNET

-10.083

8.065

-1.250

0.211

1.000

FS + RF + EFNET-SVM + VGG

13.500

8.065

1.674

0.094

1.000

FS + RF + EFNET-RF + VGG

16.917

8.065

2.098

0.036

1.000

FS + RF + EFNET-FS + SVM + VGG

-19.167

8.065

-2.377

0.017

0.489

FS + RF + EFNET-FS + RF + VGG

24.083

8.065

2.986

0.003

0.079

SVM + EFNET-FS + SVM + EFNET

-5.333

8.065

-0.661

0.508

1.000

SVM + EFNET-SVM + VGG

8.750

8.065

1.085

0.278

1.000

SVM + EFNET-RF + VGG

12.167

8.065

1.509

0.131

1.000

SVM + EFNET-FS + SVM + VGG

-14.417

8.065

-1.788

0.074

1.000

SVM + EFNET-FS + RF + VGG

-19.333

8.065

-2.397

0.017

0.463

FS + SVM + EFNET-SVM + VGG

3.417

8.065

0.424

0.672

1.000

FS + SVM + EFNET-RF + VGG

6.833

8.065

0.847

0.397

1.000

FS + SVM + EFNET-FS + SVM + VGG

9.083

8.065

1.126

0.26

1.000

FS + SVM + EFNET-FS + RF + VGG

14.000

8.065

1.736

0.083

1.000

SVM + VGG-RF + VGG

3.417

8.065

0.424

0.672

1.000

SVM + VGG-FS + SVM + VGG

-5.667

8.065

-0.703

0.482

1.000

SVM + VGG-FS + RF + VGG

-10.583

8.065

-1.312

0.189

1.000

RF + VGG-FS + SVM + VGG

-2.250

8.065

-0.279

0.78

1.000

RF + VGG-FS + RF + VGG

-7.167

8.065

-0.889

0.374

1.000

FS + SVM + VGG-FS + RF + VGG

4.917

8.065

0.610

0.542

1.000

  1. Each row tests the null hypothesis that the Sample 1 and Sample 2 distributions are the same
  2. Asymptotic significances (2-sided tests) are displayed. The significance level is 0.050
  3. a Significance values have been adjusted by the Bonferroni correction for multiple tests
  4. *Significant different p < 0.05