Skip to main content

Table 4 Performance of the ensemble model without undergoing probability calibration

From: Applying probability calibration to ensemble methods to predict 2-year mortality in patients with DLBCL

 

AUC

ECE

MCE

P

NB

0.804 (0.773–0.839)

14.206 (11.646–16.880)

38.900 (31.675–46.575)

< 0.001(< 0.001- < 0.001)

Logit

0.803 (0.773–0.835)

9.517 (7.909–11.415)

24.400 (18.875–31.900)

0.226 (0.055–0.486)

RF

0.800 (0.769–0.828)

13.569 (11.453–15.771)

36.000 (30.580–41.630)

< 0.001(< 0.001- < 0.001)

SVM

0.792 (0.762–0.821)

13.225 (11.390–15.112)

32.100 (25.420–39.970)

0.014 (0.001–0.102)

FNN

0.813 (0.780–0.845)

9.211 (7.075–10.391)

23.500 (17.675–28.925)

0.329 (0.113–0.585)

SA-EN

0.812 (0.778–0.843)

9.695 (7.968–11.699)

26.100 (19.600–32.650)

0.130 (0.025–0.332)

ECE-EN

0.813 (0.778–0.844)

9.228 (7.382–11.307)

24.500 (18.750–30.450)

0.186 (0.040–0.458)

MCE-EN

0.812 (0.777–0.843)

9.317 (7.456–11.156)

24.200 (18.700–30.525)

0.204 (0.046–0.445)

Stacking-EN

0.806 (0.771–0.834)

9.866 (8.416–11.763)

24.850 (19.275–30.425)

0.225 (0.074–0.435)

  1. SA-EN, ECE-EN, MCE-EN, and Stacking-EN represent the ensemble models obtained by combining NB, Logit, RF, SVM, and FNN models using simple averaging, weighted averaging by the ECE, weighted averaging by the MCE, and stacking method respectively
  2. In each cell M (P25 - P75): M is the median, P25 is the 25th percentile and P75 is the 75th percentile of 300 performances. The best performance in each column is in bold. The secondary best performance in each column is underlined
  3. NB Naïve Bayes, Logit Logistic regression, RF Random forest, SVM Support vector Machine, FNN Feedforward neural network
Back to article page

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

Contact us