Neural networks have an incredibly large range of applications. Classifying data is a prominent one, and this chapter is devoted to it.
5.1 Training a Network
In the previous chapter, we saw that we can obtain a trained neural network to express the XOR logical gate. In particular, we saw the following script:
n := NNetwork new.
n configure: 2 hidden: 3 nbOfOutputs: 1.
20000 timesRepeat: [
n train: #(0 0) desiredOutputs: #(0).
n train: #(0 1) desiredOutputs: #(1).
n train: #(1 0) desiredOutputs: #(1).
n train: #(1 1) desiredOutputs: #(0).
].
After evaluating ...
