A Hybrid Intelligent System Framework for the Prediction of Heart Disease Using Machine Learning Algorithms

Table 9

10-fold CV classification performance of different classifiers on selected features by LASSO FS algorithm when n = 6.

Predictive model	Classifiers performance evaluation metrics
	Turning parameters	Accuracy (%)	Specificity (%)	Sensitivity (%)	MCC	AUC (%)	Processing time (s)
Logistic regression	C = 1	85	94	74	84	86	0.012
	C = 10	87	97	76	87	88	0.019
	C = 0.1	83	90	75	84	84	0.069
K-nearest neighbor	K = 1	85	94	74	84	85	0.024
	K = 3	84	94	72	85	83	0.016
	K = 7	81	88	73	84	80	1.799
Artificial neural network	16	86	94	77	85	85	7.650
	20	82	94	70	82	81	7.362
	40	71	88	38	69	69	7.400
SVM (kernel = RBF)	C = 10, = 0.0001	85	94	74	85	84	0.019
	C = 100, = 0.001	88	96	75	88	89	0.009
SVM (kernel = linear)	C = 10, = 0.0001	84	96	74	85	85	0.023
	C = 100, = 0.0001	82	96	75	84	84	0.005
Naive Bayes	—	83	88	78	82	82	6.591
Decision tree	100	84	92	73	83	84	2.606
	50	83	90	70	83	83	2.774
Random forest	100	83	92	72	82	83	0.017