Table 2 AUC results for machine-learning methods, averaged over 100 random seeds (1 standard error).

From: A potential biomarker for treatment stratification in psychosis: evaluation of an [18F] FDOPA PET imaging approach

 

SVM linear (mean ± SD)

SVM rbf (mean ± SD)

Random Forest (mean ± SD)

Linear

Knn (2n)

Knn (3n)

GP

Dataset1

0.71 ± 0.001

0.45 ± 0.025

0.70 ± 0.008

0.74

0.63

0.68

0.64

Dataset2

0.88 ± 0.001

0.80 ± 0.007

0.83 ± 0.006

0.89

0.75

0.79

0.83

Both datasets

0.89 ± 0.001

0.80 ± 0.001

0.76 ± 0.004

0.87

0.74

0.74

0.81

  1. SVM Support Vector Machine, SVM rbf Support Vector Machine with radial basis function, Kernels. Knn (2n) K-nearest with 2 neighbours, Knn (3n) K-nearest with 3 neighbours, GP Gaussian Process.
Back to article page