Skip to main content

Table 4 Statistical results by performing PCA after REDISC and RELIC with different parameters on the Leukemia data set

From: Dimension reduction with redundant gene elimination for tumor classification

 

#genes

Sensitivity

Specificity

BACC

Precision

PPV

NPV

Correction

RELIC 0.1

18.1

0.9395

0.7433

0.8414

0.8818

0.8818

0.8758

0.8702

RELIC 0.2

55.86

0.9115

0.7783

0.8449

0.9024

0.9024

0.8007

0.8646

RELIC 0.3

205.48

0.9440

0.7700

0.8570

0.8960

0.8960

0.8692

0.8827

RELIC 0.4

790.41

0.9585

0.7983

0.8784

0.9105

0.9105

0.8975

0.9027

RELIC 0.5

2168.33

0.9735

0.8783

0.9259

0.9441

0.9441

0.9617

0.9405

RELIC 0.6

3859.52

0.9865

0.9400

0.9632

0.9716

0.9716

0.9825

0.9700

RELIC 0.7

5394.2

0.9910

0.9567

0.9738

0.9786

0.9786

0.9883

0.9784

RELIC 0.8

6545.28

0.9910

0.9633

0.9772

0.9815

0.9815

0.9875

0.9807

RELIC 0.9

7035.99

0.9955

0.9633

0.9794

0.9815

0.9815

0.9933

0.9836

REDISC 0.1

3.31

0.9865

0.6267

0.8066

0.8380

0.8380

0.9167

0.8579

REDISC 0.2

4.29

0.9805

0.6583

0.8194

0.8550

0.8550

0.8858

0.8671

REDISC 0.3

6.88

0.9695

0.7733

0.8714

0.9013

0.9013

0.9108

0.9000

REDISC 0.4

14.83

0.9655

0.8700

0.9178

0.9385

0.9385

0.9508

0.9304

REDISC 0.5

46.56

0.9735

0.9417

0.9576

0.9715

0.9715

0.9642

0.9614

REDISC 0.6

131.23

0.9690

0.9467

0.9578

0.9737

0.9737

0.9592

0.9611

REDISC 0.7

531.33

0.9795

0.9517

0.9656

0.9765

0.9765

0.9742

0.9698

REDISC 0.8

2239.81

0.9880

0.9600

0.9740

0.9798

0.9798

0.9867

0.9782

REDISC 0.9

4195.03

0.9980

0.9550

0.9765

0.9787

0.9787

0.9967

0.9823

Full Set

7129

0.9955

0.9633

0.9794

0.9815

0.9815

0.9933

0.9836