There is a great way that you can use deep learning by creating Convolutional Neural Network. Building CNN with the help of the Keras library is very simple and convenient. The core features of the CNN model are as follows:
- The input layer of CNN consists of (1, 8, 28) values.
- The first layer is called Conv2D which consists of 32 filters.
- The second layer also called Conv2D that consists of 64 filters.
- The third layer of CNN has a pool size of (2, 2).
- The fourth layer of CNN is called as Flatten used to flatten the inputs into a single dimension.
- The fifth layer of CNN is called as Dense which consists of 128 neurons.
- The sixth layer is called as Dropout which has 0.5% of its value.
- The other remaining layer consists of neurons and an activation function called ‘softmax’.
