Table 3 Overview of input Parameter grid

Random Forest

\(\bullet\) ’max_depth’: 10, 150, 500, 1000,

\(\bullet\) ’max_features’: 30, 500, 3000,

\(\bullet\) ’min_samples_leaf’: 1, 10, 100,

\(\bullet\) ’min_samples_split’: 2, 10, 100,

\(\bullet\) ’n_estimators’: 10, 100

Linear Support Vector Machine

\(\bullet\) ’loss’: ’hinge’,

\(\bullet\) ’penalty’: ’l2’,

\(\bullet\) ’alpha’: 1e-3,

\(\bullet\) ’random_state’: 42,

\(\bullet\) ’max_iter’: 5,

\(\bullet\) ’tol’: None

Logistic Regression

\(\bullet\) ’n_jobs’: 1,

\(\bullet\) ’C’: 1

Neural Network (BERT)

\(\bullet\) ’batch_size’: 3,

\(\bullet\) ’lr’: 1e-5,

\(\bullet\) ’eps’: 1e-8,

\(\bullet\) ’epochs’: 5,

\(\bullet\) ’seed_val’: 17