Table 10 Examined hyperparameters for proposed ML models
From: Application of machine learning techniques for predicting survival in ovarian cancer
Model | Hyperparameters |
|---|---|
KNN | algoritgm: (kd_tree, ball_tree, auto) p: (1, 2) n_neighbors: (1–15) other hyper-parameters values: default |
SVM | kernel: (rbf, poly, linear) gamma: (0.01, 0.1, 1, 10, 50, 100) C: (0.01, 0.1, 1, 10, 50, 100) other hyper-parameters values: default |
DT | spliter: (best, random) max_depth: (5, 10, 20, None) criterion: (entropy, gini) other hyper-parameters values: default |
RF | criterion: (entropy, gini) max_depth: (5, 10, 15, None) n_estimators: (50, 100, 150, 200) other hyper-parameters values: default |
AdaBoost | n_estimators: (50, 100, 150, 200) learning_rate: (0.5–2.0) algorithm: (SAMME, SAMME.R) other hyper-parameters values: default |
XGBoost | sampling method: (gradiant_based, uniform, subsample) eta: (0.1–0.9) booster: (dart, gblinear, gbtree) other hyper-parameters values: default |