Table 2.
RNA Expression Classifier Performance
| Datasets | Classifiers: | RF | WSRF | CIRF |
|---|---|---|---|---|
| AUCs | ||||
| Full RNA | 0.632 ± 0.106 | 0.596 ± 0.038 | 0.629 ± 0.105 | |
| SPCA | 0.640 ± 0.128 | 0.626 ± 0.114 | 0.658 ± 0.044 | |
| Nested CV Runtimes | – | – | – | |
| Full RNA | 52 h | 185 h | 85 h | |
| SPCA | 12 min | 1.9 h | 30 min | |
AUC and approximate runtime values for the RNA expression feature sets. The best value in each row is bolded. Here, runtimes are evaluation times for a given classifier on a given feature set via 10-fold nested cross validation with the internal cross validation procedures as described in Methods. Computations performed on the University of Iowa’s Argon High-Performance Computing cluster