[Retracted] An AI-Driven Hybrid Framework for Intrusion Detection in IoT-Enabled E-Health

Table 3

10-Fold results of Cu-LSTM+ GRU, Cu-GRU+ DNN, and Cu-BLSTM.

	Model	1	2	3	4	5	6	7	8	9	10
Accuracy (%)	Cu-LSTM+GRU	98.25	98.23	99.15	98.89	99.08	99.31	99.16	99.12	99.16	99.84
	Cu-GRU+DNN	97.56	97.21	97.86	97.54	98.54	98.57	99.15	98.81	98.62	98.86
	Cu-BLSTM	98.36	98.36	98.41	98.93	98.87	98.87	98.69	98.36	98.24	98.29
F1-score (%)	Cu-LSTM+GRU	98.24	98.63	99.68	99.06	99.06	99.25	99.19	99.34	99.08	99.68
	Cu-GRU+DNN	98.62	98.45	98.15	98.62	98.62	98.74	99.11	99.15	98.82	98.18
	Cu-BLSTM	98.94	98.91	98.29	98.29	98.68	98.15	98.19	98.81	99.16	99.43
Recall (%)	Cu-LSTM+GRU	98.96	98.92	99.26	98.61	98.21	98.61	98.89	98.97	98.69	98.92
	Cu-GRU+DNN	98.15	98.06	98.04	98.04	98.61	98.25	98.54	98.95	99.15	98.87
	Cu-BLSTM	98.15	98.16	98.16	98.85	98.71	98.06	98.15	98.64	98.64	98.86
Precision (%)	Cu-LSTM+GRU	98.16	98.68	99.14	99.14	99.32	99.36	99.86	99.51	98.91	98.34
	Cu-GRU+DNN	98.69	98.85	98.85	98.09	97.93	97.19	98.14	98.31	98.16	98.31
	Cu-BLSTM	98.19	98.96	98.48	98.48	98.86	98.46	98.69	99.05	99.17	98.78