Hands-On Artificial Intelligence for Beginners

Setting up network parameters in our MNIST example

Looking at our MNIST example, we can now set up our overall network parameters. We'll define input_layer_size as the size of the incoming data, 784. This will give us the number of neurons in the input layer. The parameters hidden_one and hidden_two will define the amount of neurons in the hidden layer. Meanwhile, number_classes constructs the output layer as the potential number of classes that a piece of input data could be classified as:

## Size of the input datainput_layer_size = 784 ## Define the size of the hidden layers; We want them to be smaller than the inputhidden_layer_one = 256hidden_layer_two = 256 ## Size of the potential output classesnumber_classes = 10 

In total, these will ...