Table 3 Results of the CPH model.

From: Explainable machine learning can outperform Cox regression predictions and provide insights in breast cancer survival

Feature

HR

95% CI

z

p

age

1.055

1.054

1.057

80.191

0.000

ratly

1.604

1.434

1.794

8.269

0.000

rly

0.999

0.996

1.002

-0.672

0.501

ptmm

1.008

1.007

1.010

13.604

0.000

ply

1.021

1.012

1.031

4.456

0.000

pts

I

1.000

IIA

1.096

1.045

1.149

3.801

0.000

IIB

1.352

1.265

1.446

8.804

0.000

IIIA

1.579

1.448

1.722

10.344

0.000

IIIB

2.250

2.034

2.490

15.707

0.000

IIIC

1.862

1.614

2.148

8.526

0.000

grd

1

1.000

2

1.132

1.081

1.187

5.213

0.000

3

1.368

1.296

1.443

11.472

0.000

mor

Ductal

1.000

Lobular

0.956

0.908

1.007

1.680

0.093

Mixed

1.043

0.961

1.133

1.843

0.313

Other

0.927

0.854

1.006

-0.655

0.071

rec

 

Triple+

1.000

HR+

1.006

0.930

1.089

0.150

0.881

HR–

1.168

1.055

1.292

3.005

0.003

Triple–

1.473

1.348

1.609

8.577

0.000

  1. HR hazard ration, CI confidence interval, HR + hormone receptor positive, HR - hormone receptor negative.