Cross-Subject Seizure Detection in EEGs Using Deep Transfer Learning

Table 2

The metrics of each model with different full connection size.

Size of FC	Base model	Loss	Accuracy	Recall	Precision	Mcor
	Model-1	0.0631	0.9837	0.9837	0.9839	0.9674
	Model-2	0.0698	0.9816	0.9844	0.9789	0.9632
	Model-3	0.1790	0.9578	0.9596	0.9563	0.9157
	Model-1	0.0638	0.9826	0.9844	0.9810	0.9653
	Model-2	0.0731	0.9805	0.9823	0.9788	0.9611
	Model-3	0.1739	0.9663	0.9806	0.9526	0.9331
	Model-1	0.0596	0.9833	0.9823	0.9844	0.9667
	Model-2	0.0712	0.9791	0.9851	0.9734	0.9583
	Model-3	0.1820	0.9568	0.9724	0.9430	0.9140
	Model-1	0.0618	0.9830	0.9851	0.9810	0.9660
	Model-2	0.0690	0.9815	0.9844	0.9789	0.9631
	Model-3	0.2020	0.9567	0.9752	0.9406	0.9142
	Model-1	0.0637	0.9816	0.9851	0.9782	0.9631
	Model-2	0.0672	0.9808	0.9851	0.9768	0.9617
	Model-3	0.2091	0.9564	0.9752	0.9400	0.9135
	Model-1	0.0673	0.9809	0.9837	0.9782	0.9618
	Model-2	0.0758	0.9777	0.9872	0.9687	0.9555
	Model-3	0.1649	0.9642	0.9851	0.9456	0.9292
	Model-1	0.0624	0.9844	0.9823	0.9865	0.9688
	Model-2	0.0664	0.9808	0.9872	0.9748	0.9618
	Model-3	0.1452	0.9610	0.9653	0.9572	0.9221
	Model-1	0.0678	0.9769	0.9851	0.9693	0.9541
	Model-2	0.0669	0.9815	0.9851	0.9782	0.9632
	Model-3	0.1201	0.9773	0.9879	0.9674	0.9548
	Model-1	0.0653	0.9794	0.9844	0.9747	0.9589
	Model-2	0.0841	0.9720	0.9880	0.9575	0.9445
	Model-3	0.1198	0.9745	0.9844	0.9653	0.9491