Hands-On Artificial Intelligence for Beginners

Putting it all together

Now that we've gone through all of the elements of a basic feedforward neural network, it's time to begin assembling the pieces. The first thing we'll do is define our hyperparameters. We'll train the network for 50 epochs, each time feeding a batch of 100 samples into the network. At the end of each epoch, the batch_size parameters will tell the network to print out the current value of the loss, as well as the accuracy of the network:

## Network Parametersepochs = 15batch_size = 100display_step = 1

Next, we'll create placeholders for our MNIST data features (x) and their correct label (y), which we'll use to represent the data while constructing the network:

 # Create the Variablesx = tf.placeholder("float", [None, ...