Table 3 Performance evaluation of the... | Emerald Publishing

Table 3

Performance evaluation of the ResNet50-CBAM

Data split	Train test split			Five-fold cross-validation
Data split	Train test split			Model A		Model B
Metric	Model A (%)	Model B (%)	Average (%)	Mean (%)	Standard deviation	Mean (%)	Standard deviation
Accuracy	99.43	98.50	98.97	99.35	0.009	98.68	0.014
Recall	99.01	96.11	97.56	98.54	0.022	96.49	0.021
Precision	98.7	97.74	98.22	98.90	0.014	97.68	0.017
F1-Score	99.0	96.91	97.96	98.70	0.018	97.04	0.014
AUC	99.25	97.63	98.44	99.06	0.013	97.75	0.015

Data split	Train test split			Five-fold cross-validation
Data split	Train test split			Model A		Model B
Metric	Model A (%)	Model B (%)	Average (%)	Mean (%)	Standard deviation	Mean (%)	Standard deviation
Accuracy	99.43	98.50	98.97	99.35	0.009	98.68	0.014
Recall	99.01	96.11	97.56	98.54	0.022	96.49	0.021
Precision	98.7	97.74	98.22	98.90	0.014	97.68	0.017
F1-Score	99.0	96.91	97.96	98.70	0.018	97.04	0.014
AUC	99.25	97.63	98.44	99.06	0.013	97.75	0.015

Source(s): Table created by the authors