[Retracted] Performance Analysis of Deep Learning Models for Binary Classification of Cancer Gene Expression Data

Table 5

Architectural parameter setting for 2DCNN.

Dataset	Dense layer size	No. of filters, filter size	Train loss	Validation loss
Colon (ANOVA)	50	64 (9,9)	9.8e − 10	15.26
Colon (IG)	20	32 (5,5)	0.112	6.42
Pancreatic (ANOVA)	40	128 (5,5)	2.45e − 6	10.34
Pancreatic (IG)	40	128 (5,5)	1.08e − 9	5.24
Breast (ANOVA)	100	64 (5,5)	6.9e − 1	0.67
Breast (IG)	50	64 (5,5)	2.23e − 5	7.55
Lung (ANOVA)	50	64 (7,7)	2.11e − 7	10.1
Lung (IG)	40	64 (7,7)	1.23e − 11	6.34