Table 1 Breast cancer survival prognosis researches using SEER data
Sources | Class distribution | Classifier methods | Accuracy performances |
|---|---|---|---|
Delen et al. [8] | Survival: 46% | C5 DT | 93.62% |
Non-survival: 54% | ANN | 91.21% | |
LR | 89.20% | ||
Bellaachia and Guven [9] | Survival: 76.80% | C4.5 DT | 86.70% |
Non-survival: 23.20% | ANN | 86.50% | |
Naïve BN | 84.50% | ||
Endo et al. [10] | Survival: 81.50% | LR | 85.80% |
Non-survival: 18.50% | J48 DT | 85.60% | |
DT (with naïve Bayes) | 84.20% | ||
ANN | 84.50% | ||
Naïve BN | 83.90% | ||
BN | 83.90% | ||
ID3 DT | 82.30% | ||
Liu et al. [11] | Survival: 86.52% | C5 DT | 88.05% (AUC = 0.607) |
Non-survival: 13.48% | Under-sampling + C5 DT | 74.22% (AUC = 0.748) | |
|  |  | Bagging algorithm + C5 DT | 76.59% (AUC = 0.768) |