Table 2.

Optimum hyperparameters for SVM and RF models through grid search approach.

Model	Optimum Hyperparameters 1
SVM	kernel = ‘rbf’; C = 10; gamma = 0.01
RF	n_estimators = 300; max_depth = none; optimal number of features = 1/3 of total features
MLP	Optimization algorithm = stochastic gradient descent; epochs = 50; learning rate = 0.01; neurons for the first hidden layer = 96; neurons for the second hidden layer = 64
CNN	Optimization algorithm = stochastic gradient descent; epochs = 50; learning rate = 0.01; neurons for the first hidden layer = 32; neurons for the second hidden layer = 16

¹ The optimal hyperparameters detected through grid search.