Table 6.

Result of applying models with full features and selected features for the Cleveland dataset.

Approaches	Models	Features	Matrix Performance
			ACC	PRE	REC	F1
Regular ML approach	RF	Full features	86.34	86.34	86.34	86.34
		Selected features	82.93	82.99	82.93	82.91
	LR	Full features	67.32	67.43	67.3	67.18
		Selected features	73.17	73.19	73.17	73.14
	DT	Full features	82.44	82.46	82.44	82.44
		Selected features	81.95	82.01	81.95	81.93
	NB	Full features	60.00	60.05	60.00	59.74
		Selected features	64.88	64.90	64.88	64.88
	KNN	Full features	60.00	60.25	60.00	59.92
		Selected features	66.34	66.62	66.34	66.29
The hybrid models	CNN-LSTM	Full features	89.76	89.96	89.76	89.75
		Selected features	86.34	86.41	86.34	86.34
	CNN-GRU	Full features	88.29	89.06	88.29	88.26
		Selected features	85.85	86.92	85.85	85.78
The proposed model	Stacking SVM	Full features	97.17	97.42	97.17	97.15
		Selected features	91.22	91.29	91.22	91.22