Table 1 Cross validation-based performance of the SVM model on TCGA testing dataset
From: Deep learning algorithm reveals two prognostic subtypes in patients with gliomas
Ten folds cross validation | C-index, mean (SD) | Brier score, mean (SD) | Log-rank p value, geo.mean |
|---|---|---|---|
TCGA 2-omics testing dataset (40%) | 0.92 (0.02) | 0.16 (0.02) | 4.68E−12 |
TCGA mRNA testing dataset | 0.92 (0.02) | 0.17 (0.02) | 3.73E−12 |
TCGA methylation testing dataset | 0.95 (0.02) | 0.16 (0.03) | 6.33E−13 |