Shallow Neural Network Plotter

Add bias/weight-colored diagram to Deep Learning Toolbox training interface for shallow neural networks.

A function designed to simply plot a neural network architecture consisting of a couple of layers. Lines correspond to weights and nodes correspond to biases which are colored based on relative magnitude to the maximum and minimum weight or bias present in the plot.
Provides integration with the Deep Learning Toolbox's network training interface, and adds a button to bring up the network plot which will update at each epoch.

Example Use:
network = fitnet(num_hidden_layers);
network.plotFcns = [network.plotFcns 'plotShallowNetwork'];
[t_network, t_Record] = train(network, in_data, out_data);