Table 5.

Results of comparison of different approaches.

Model no.	Approaches	Accuracy (%)	F1-score (%)
1	Gaussian-Kernel principal components Analysis (PCA)+CNN33	93.90	–
2	DT classifier + binary grey wolf optimization (BGWO)32	93.95	–
3	Adam + maximum entropy Markov model (MEMM)34	90.91	–
4	CNN-pff model36	91.94	–
5	Hybrid model between LSTM-CNN35	95.56	–
6	Semisupervised deep model37	94.05	–
7	GAF+ DenseNet169 (the proposed model)	97.83	97.83