Machine learning models and hyperparameter settings
| Method | Class (Package) | Parameter |
|---|---|---|
| SVM | SVC (scikit-learn) | C: 1.0, kernel: rbf, degree: 3, gamma: scale, coef: 0.0, shrinking: True, probability: True, tol: 0.001, class weight: False, max_iter: 1, decision function shape: ovr, break ties: False |
| LR | LogisticRegression (scikit-learn) | penalty: l2, dual: False, tol: 0.0001, C: 1.0, fit intercept: True, intercept scaling: 1, class weight: None, solver: lbfgs, max iter: 100, multi class: auto, wart start: False |
| XGB | XGBClassifier (xgboost) | booster: gbtree, learning rate: 0.3, gamma: 0, max depth: 6, min child weight: 1, max delta step: 0, subsample: 1, sampling method: uniform, colsample bytree: 1, colsample bylevel: 1, colsample bynode: 1, lambda: 1, alpha: 0, tree method: auto, scale pos weight: 1, refresh leaf: 1, max leaves: 0, max bin: 256, num parallel tree: 1 |
| RF | RandomForestClassifier (scikit-learn) |
n estimator: 100, criterion: gini, max depth: None, min samples split: 2, min samples leaf: 1, min weight fraction leaf: 0, max features: sqrt, max leaf nodes: None, min impurity decrease: 0, bootstrap: True, oob score: False, warm strat: False, class weight: None, ccp alpha: 0, max samples: None, monotonic: None |
| DNN | Model (TensorFlow) | k: 5, input shape: 2048, layers: 2048, 1024, 512, 256, train size: 0.6, validation size: 0.2, test size: 0.2l2 regularization: null, batch normalization: False, activation function: relu, loss function: BinaryCrossentropy, learning rate: 0.001, optimizer: Adam, metric: BinaryAccuracy, AUC, early stop monitor: val loss, early stop patience: 10, class weight: False, batch size: 256, epochs: 1000, seed: 42 |
SVM, Support Vector Machine; LR, Logistic Regression; XGB, XGBoost; RF, Random Forest; DNN, Deep Neural Network.