Skip to main content

Table 3 Random Forest models for overall survival prediction. C-index results are presented for each combination of variables in the training and test cohorts

From: Improved personalized survival prediction of patients with diffuse large B-cell Lymphoma using gene expression profiling

  Training Cohort Test Cohort
GEP_0.01 0.5934 0.6301
GEP_0.05 0.7530 0.6649
GEP_0.1 0.7783 0.7415
Age, Gender, Stage 0.6340 0.6202
Age, Gender, Stage, COO 0.6761 0.6837
Age, Gender, Stage, 4-gene expression cluster 0.6725 0.6971
Age, Gender, Stage, COO, 4-gene expression cluster 0.7059 0.7221
GEP_0.1, 4-gene expression cluster 0.7792 0.7558
GEP_0.1, COO 0.7784 0.7487
Age, Gender, Stage, GEP_0.1 0.7788 0.7522
Age, Gender, Stage, GEP_0.1, 4-gene expression cluster 0.7889 0.7416
Age, Gender, Stage, GEP_0.1, COO 0.7854 0.7538
Age, Gender, Stage, COO, GEP_0.1, 4-gene expression cluster 0.7896 0.7596
Age, Gender, Stage, COO, GEP_0.1, 4-gene expression cluster
(parameter optimized)
0.8051 0.7615
Age, Stage, COO, 4-gene expression cluster, 50 genes
(variable reduction, parameter optimization)
0.8404 0.7942