Table 20.

Results with standard deviation of predictor classification using h2o in TCGA.

Algorithm	AUC	ACC	Average ACC per class	MSE	Average MSE per class
Standard variables and without replication
DL	0.591 ± 0.019	0.706 ± 0.076	0.639 ± 0.024	0.147 ± 0.032	0.361 ± 0.024
XGBoost	0.631 ± 0.029	0.695 ± 0.101	0.634 ± 0.038	0.091 ± 0.012	0.366 ± 0.037
GBM	0.647 ± 0.046	0.826 ± 0.038	0.639 ± 0.035	0.109 ± 0.017	0.361 ± 0.035
RF	0.642 ± 0.046	0.806 ± 0.081	0.629 ± 0.036	0.090 ± 0.014	0.371 ± 0.036
Replicated and unbalanced TCGA
DL	0.961 ± 0.009	0.925 ± 0.004	0.883 ± 0.010	0.064 ± 0.005	0.117 ± 0.010
XGBoost	0.962 ± 0.005	0.929 ± 0.004	0.897 ± 0.009	0.055 ± 0.003	0.102 ± 0.009
GBM	0.961 ± 0.005	0.929 ± 0.006	0.892 ± 0.012	0.056 ± 0.002	0.108 ± 0.012
RF	0.959 ± 0.005	0.926 ± 0.004	0.882 ± 0.003	0.061 ± 0.002	0.118 ± 0.003
GLM	0.933 ± 0.007	0.890 ± 0.018	0.859 ± 0.005	0.109 ± 0.006	0.143 ± 0.013
Replicated and balanced TCGA
DL	0.961 ± 0.009	0.925 ± 0.004	0.882 ± 0.010	0.064 ± 0.005	0.117 ± 0.010
XGBoost	0.961 ± 0.005	0.929 ± 0.004	0.897 ± 0.009	0.056 ± 0.003	0.108 ± 0.012
GBM	0.962 ± 0.005	0.929 ± 0.006	0.892 ± 0.012	0.056 ± 0.002	0.108 ± 0.012
RF	0.959 ± 0.005	0.926 ± 0.004	0.882 ± 0.003	0.061 ± 0.002	0.118 ± 0.003
GLM	0.933 ± 0.007	0.890 ± 0.010	0.859 ± 0.005	0.019 ± 0.006	0.143 ± 0.013