Table 3.

Summary of parameters used in each classifier.

Classifier	Parameters
Random forest	bag size percent = 100, batch size = 100, number of execution slots = 1, max depth = 0 (unlimited), number of randomly chosen attributes = 0, number of iterations to be performed = 100, minimum number of instances per leaf = 1.0, minimum variance for split = 0.001, random number seed to be used = 1
MLP	learningrate = 0.3, momentum = 0.2, number of epochs used for training = 500, validation set size = 0 (the network will by training for the specified number of epochs), seed = 0, validation threshold = 20, hidden layers = ((number of attributes + classes)/2)
SVM	C = 1.0, kernel = radial basis function (RBF), degree = 3, gamma = scale, shrinking = true, probability = false, tol = 0.001, cache size = 200, max iter = −1, random state = none
XGBoost	max depth = 7, learning rate = 0.1, ite = 1000, gama = 0, max delta step = 1, objective = “multi:softmax”