Research Article
Comparative Analysis of Skin Cancer (Benign vs. Malignant) Detection Using Convolutional Neural Networks
Table 9
Layers and parameters of sequential model 3.
| | Type of layers | Outcome structure | Number of parameters |
| | Conv-1 (2d) | (Nil, 222 × 222 × 64) | 1792 | | Pooling-1 (max) | (Nil, 111 × 111 × 64) | Null | | Normalization-1 batch | (Nil, 111 × 111 × 64) | 256 | | Layer dropout-1 | (Nil, 111 × 111 × 64) | Null | | Conv-2 (2d) | (Nil, 109 × 109 × 64) | 36928 | | Pooling-2 (max) | (Nil, 52 × 52 × 64) | Null | | Conv-3 (2d) | (Nil, 52 × 52 × 64) | 36928 | | Pooling-3 (max) | (Nil, 26 × 26 × 64) | Null | | Layer flatten-1 | (Nil, 43264) | Null | | Layer dense-1 | (Nil, 128) | 5537920 | | Normalization-2 batch | (Nil, 128) | 512 | | Layer dropout-2 | (Nil, 128) | Null | | Layer dense-2 | (Nil, 1) | 129 |
| | Total number of parameters: 5,614,465; number of trainable parameters: 5,614,082; number of nontrainable parameters: 384 |
|
|
