Comparative Analysis of Skin Cancer (Benign vs. Malignant) Detection Using Convolutional Neural Networks

Table 5

Layers and parameters of sequential model 1.

Type of layers	Outcome structure	Number of parameters
Conv-1 (2d)	(Nil, 222 × 222 × 50)	1400
Pooling-1 (max)	(Nil, 111 × 111 × 50)	Null
Layer Dropout-1	(Nil, 111 × 111 × 50)	Null
Conv_2 (2d)	(Nil, 109 × 109 × 20)	9020
Pooling-2 (max)	(Nil, 54 × 54 × 20)	Null
Layer Dropout-2	(Nil, 54 × 54 × 20)	Null
Conv_3 (2d)	(Nil, 54 × 54 × 20)	3620
Pooling-3 (max)	(Nil, 26 × 26 × 20)	Null
Layer Dropout-3	(Nil, 26 × 26 × 20)	Null
Layer flatten	(Nil, 13520)	Null
Layer dense	(Nil, 100)	1352100
Layer dropout-4	(Nil, 100)	Null
Layer dense-1	(Nil, 50)	5050
Layer dropout-5	(Nil, 50)	Null
Layer dense-2	(Nil, 1)	51
Total number of parameters: 1,371,241;
number of trainable parameters: 1,37,241;
number of nontrainable parameters: null