Table 1.

Performance comparison of hybrid models.

Approach	Classifier	Accuracy (%)	Precision (%)	Recall (%)	F1 Score (%)
Single model	SVM	81	93	66	77
	DT	72	77	62	69
	RF	81	92	68	78
	NB	63	65	63	61
	CNN	92	93	86	89
	CapsNet	82.9	86	78	81.8
	Wide ANN	78	88	69	77.3
Hybrid model	DT + CNN	89	89	88.5	88.5
	RF + CNN	90	91.5	90	90
	NB + CNN	86	86	86	86
	SVM + CNN (Proposed)	94	95	94.5	94.5