Skip to main content

Table 3 Results obtained for the selected subset SEM (semantic) LH (left hemisphere) BA 21 and BA47 in terms of AUC as the performance metric for each iteration of the outer MCCV, using the XGBoost algorithm (n_estimators = 1200, learning rate = 0.01, subsample = 0.7, max_depth = 3)

From: Machine learning–XGBoost analysis of language networks to classify patients with epilepsy

Iteration number

1

2

3

4

5

6

7

8

9

10

11

12

Subset selected

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

SEM L21 L47

AUC (%)

93.75

87.50

87.50

93.75

93.75

93.75

100

83.33

83.33

93.75

87.50

100