Table 2.

The optimal hyperparameters of machine learning classifiers.

Models	Hyperparameters
Decision tree	{Max depth: 3, max leaf nodes: 4, min samples leaf: 5, and min samples split: 165}
K-neighbors	{n neighbors: 30}
XgBoost	{Learning rate: 0.01, max depth: 3, n estimators: 100, and subsample: 0.3}
Gradient boosting	{Learning rate: 0.05, max depth: 1, n estimators: 30, and subsample: 0.3}
Logistic regression	{C: 0.1, l1 ratio: 0.01, max iter: 10000, and solver: Liblinear}
Support vector classifier	{C: 0.5, degree: 1, kernel: “Linear”}
Light GBM	{Learning rate: 0.2, max depth: 3, n estimators: 15, and subsample: 0.3}
Random forest	{Max depth = 2, max features = 3, and n estimators = 5}
AdaBoost	{Learning rate: 0.2, n estimators: 20}
Bernoulli naïve bayes	{Default}