Cough event classification by pretrained deep neural network

Table 3 Result table with pretrained neural network on PD.

# layers	# hidden units	Sens.	Spec.	F1	Mac. Ave.	Mic. Ave.
1	512	0.804	0.906	0.678	0.855	0.892
1	1024	0.793	0.917	0.691	0.855	0.899
1	1536	0.799	0.915	0.692	0.857	0.899
1	2048	0.794	0.917	0.693	0.856	0.9
2	512	0.793	0.914	0.685	0.853	0.896
2	1024	0.794	0.919	0.695	0.857	0.901
2	1536	0.803	0.916	0.695	0.859	0.9
2	2048	0.804	0.921	0.705	0.863	0.905
3	512	0.792	0.912	0.682	0.852	0.895
3	1024	0.793	0.914	0.686	0.853	0.897
3	1536	0.811	0.911	0.691	0.861	0.897
3	2048	0.8	0.915	0.691	0.857	0.898
4	512	0.791	0.916	0.688	0.853	0.898
4	1024	0.803	0.914	0.691	0.858	0.898
4	1536	0.801	0.916	0.694	0.859	0.9
4	2048	0.782	0.921	0.693	0.852	0.902
Baseline Method		0.945	0.767	0.563	0.856	0.792

ISSN: 1472-6947