Medical Text Classification Using Hybrid Deep Learning Models with Multihead Attention

Table 9

Analysis of results with different BiGRU layers.

Layers	Dataset	Accuracy (%)	Precision (%)	Recall (%)	F1-score (%)
1	Hallmarks	67.81	59.16	57.19	58.15
	AIM	91.11	83.35	83.26	83.30
2	Hallmarks	72.11	71.87	72.75	72.30
	AIM	94.01	88.95	84.91	86.88
3	Hallmarks	71.91	69.27	65.91	67.54
	AIM	81.16	76.18	75.37	75.77
4	Hallmarks	72.73	66.17	64.98	65.56
	AIM	69.69	54.99	57.21	56.07