Comparing machine learning algorithms to predict 5-year survival in patients with chronic myeloid leukemia

Table 2 Best hyperparameters of all the trained algorithms

Num	Data mining Models	Hyper-parameters	f-score
1	Decision tree (j48)
2	MLP classifier	‘Learning rate’ = ’constant’, hidden_layer_size’ = (100,100,100), ‘alpha’ = 0.05, ‘activation’ = ’rulo’	87.6
3	SVM (kernel = linear)	C = 100, G = 0.0001	83.04
4	SVM (kernel = RBF)	C = 10, G = 0.001	81.9
5	XG Boost Classifier	‘min_chid_weigh’ = 1’max_depht’ = 12,’learning_rate’ = 0.1, ‘gamma’ = 0.4, ‘colsample_bytree’ = 0.3	81.02
6	KNN	K = 5	67.1
7	Pattern recognition network	57-10-5-2	69.02
8	Probabilistic neural network	57-2, Spread = 0.1	70.01

SVM support vector machine, XG Boost eXtreme gradient boosting, KNN K-nearest neighborhood

ISSN: 1472-6947