Table 2 ML moder parameters.
From: Explainable illicit drug abuse prediction using hematological differences
Classifier | Hyper-parameter | Value |
|---|---|---|
GBM | Estimators | 200 |
Learning rate | 0.01 | |
Max depth | 6 | |
RF | Estimators | 1000 |
Min samples leaf | 1 | |
Min samples split | 2 | |
Max depth | 4 | |
SVM | Kernel | Linear |
Regularization | C=0.1 | |
Degree | 3 | |
XGB | Estimators | 50 |
Learning rate | 0.1 | |
Max depth | 9 | |
LR | Penalty | L1 |
Solver | Liblinear | |
Max iter | 100 | |
BPNN | Hidden layer sizes | 128, 64, 32 |
Max iter | 500 | |
Activation | ReLU | |
AdaBoost | Estimators | 100 |
Learning rate | 0.5 | |
LGB | Estimators | 50 |
Num leaves | 20 | |
Learning rate | 0.05 | |
Max depth | 3 |
