Applying machine learning to high-dimensional proteomics datasets for the identification of Alzheimer’s disease biomarkers

Table 7 Hyperparameters for machine learning models using BayesSearchCV

Model	Hyperparameter	Value
XGBoost	eta max_depth n_jobs n_estimators objective num_classes	Real(0.1, 0.5) Integer(1, 20) -1 Integer(50, 500) multi:softmax 3
Logistic Regression	penalty C solver multi_class max_iter n_jobs l1_ratio	elasticnet Real(0.000001, 100) saga multinomial Integer(1000, 12000) -1 Real(0, 1)
Random Forest	n_estimators max_depth n_jobs min_samples_leaf	Integer(5, 500) Integer(2, 50) -1 Integer(1, 5)

This table lists the hyperparameter ranges for each machine learning model optimized with BayesSearchCV. Tuning options for XGBoost, Logistic Regression, and Random Forest use specific values or ranges (e.g., Real, Integer) to define search spaces, improving model performance by exploring various parameter combinations

ISSN: 2045-8118