The past few sections have introduced the reader to the neuron and the feedforward neural network. The neuron is the building block for all the network architectures that will be presented in this book. The concept of a learning system, which follows, will prepare the reader for the rest of the text. The reader may recall that the perception learning rule required the addition of a desired output before the network could adapt the weights to find the line that separated one class from the other. This process of using desired outputs for training the neural network is known as supervised training.
Online access to SPIE eBooks is limited to subscribing institutions.