Artificial Intelligence-Based Classification of Chest X-Ray Images into COVID-19 and Other Infectious Diseases

Table 2

Hyperparameters used to train CovidPred models.

(a) Original images/single augmentation-based models (for 26 models)

Sr. No.	Hyperparameter	Value
1	Number of iterations	400
2	Batch size	16
3	Number of epochs	24
4	Image size	256
5	Internal validation size	0.1
6	Filter size (1st convolutional layer)	3
7	Number of filters (1st convolutional layer)	32
8	Filter size (2nd convolutional layer)	5
9	Number of filters (2nd convolutional layer)	64
10	Filter size (3rd convolutional layer)	7
11	Number of filters (3rd convolutional layer)	128
12	Fully connected layer size	256

(b) Original and all augmentation-based models (for combined image-based models)

Sr. No.	Hyperparameter	Value
1	Number of iterations	5568/11136/23300
2	Batch size	32
3	Number of epochs	24/49/101
4	Other hyperparameters	Same as given in (a)