In neural networks, there are a number of architectures implementing unsupervised learning; however, the scope of this book will cover only the Kohonen neural network, developed in Chapter 4, Self-Organizing Maps.
Kohonen Networks, which have been covered in Chapter 4, Self-Organizing Maps are now used in a modified fashion. Kohonen can produce a shape in one or two dimensions at the output, but here we are interested in clustering, which can be reduced in only one dimension.
Tip
Actually the Kohonen neural network implemented in this framework considers the dimensions zero, one, and two, where zero means no connections between the output neurons and one means they form a line, and two means a grid. For this chapter's example, we will need a Kohonen network with no connected output neurons, therefore, the dimension will be zero.
In addition, clusters may be related or not to each other, so the vicinity of neurons can be ignored for now in this chapter, which means only one neuron will be activated and their neighbors will remain unchanged. And so, the neural network will adjust its weights to match data to an array of clusters:
The training algorithm will be the competitive learning, whereby the neuron with the greatest output has its weights adjusted. By the end of training, all the clusters of a neural network are expected to be defined. Note that there are no links between output neurons, meaning that only one input is active at the output.