Introduction

Get the entire book!

In Chapter 5 backpropagation was introduced. Backpropagation is a very effective means of training a neural network. However, there are some inherent flaws in the back propagation training algorithm. One of the most fundamental flaws is the tendency for the backpropagation training algorithm to fall into a “local minima”. A local minimum is a false optimal weight matrix that prevents the backpropagation training algorithm from seeing the true solution.

In this chapter you will see how you can use other training techniques to supplement backpropagation and elude local minima. In the previous two chapters you learned about two net optimization techniques that were used to find the minimal path that a traveling salesman might travel. This chapter will begin by introducing you to exactly what a local minimum is. Then the genetic and simulated annealing algorithms will be used to allow the backpropagation training algorithm to escape a local minima and seek a more optimal solution, if one does exist. We will begin by examining global and local minima.