Table 7.

Target class count for training and testing sets.

Class	Hyper-Parameters	Tuning Range
LR	solver = liblinear, C = 5.0	solver = {liblinear, saga, sag}, C = {1.0 to 8.0}
SVM	kernel = ‘linear’, C = 5.0	kernel = {‘linear’, ‘poly’, ‘sigmoid’} C = {1.0 to 8.0}
RF	n_estimators = 200, max_depth = 20	n_estimators = {10 to 300}, max_depth = {2 to 50}
GBM	n_estimators = 200, max_depth = 20, learning_rat = 0.5	n_estimators = {10 to 300}, max_depth = {2 to 50}, learning_rat = {0.1 to 0.9}
ADA	n_estimators = 200, max_depth = 20, learning_rat = 0.5	n_estimators = {10 to 300}, max_depth = {2 to 50}, learning_rat = {0.1 to 0.9}