Table 10 Performance comparison (precision and recall) with recent methods, based on the second study with the 450k DNA methylation profile dataset
Cancer type | Method | Precision | Recall |
|---|---|---|---|
Breast Cancer (BRCA) | DNN (Zheng et al.) | 0.981 | 1.0 |
RF (Modhukur et al.) | 0.976 | 0.932 | |
Proposed method | 0.986 | 0.986 | |
Colon Cancer (COAD) | DNN (Zheng et al.) | – | – |
RF (Modhukur et al.) | 0.993 | 0.980 | |
Proposed method | 0.988 | 0.988 | |
Kidney (KIRC) | DNN (Zheng et al.) | 1.0 | 1.0 |
RF (Modhukur et al.) | 1.0 | 0.974 | |
Proposed method | 1.0 | 0.99 | |
Liver (LIHC) | DNN (Zheng et al.) | 0.66 | 1.0 |
RF (Modhukur et al.) | 1.0 | 0.99 | |
Proposed method | 1.0 | 1.0 | |
Lung (LUSC) | DNN (Zheng et al.) | 1.0 | 0.66 |
RF (Modhukur et al.) | 0.91 | 0.953 | |
Proposed method | 1.0 | 0.986 | |
Prostate (PRAD) | DNN (Zheng et al.) | 1.0 | 1.0 |
RF (Modhukur et al.) | 1.0 | 0.980 | |
Proposed method | 1.0 | 0.980 | |
Thyroid (THCA) | DNN (Zheng et al.) | 1.0 | 0.987 |
RF (Modhukur et al.) | 1.0 | 1.0 | |
Proposed method | 0.975 | 0.99 |