In this chapter, we learned how to solve a handwritten digit-recognition problem. Starting from the basics of the OCR and computer vision concepts, we learned how to elaborate simple images.
We analyzed different types of generative models. A Boltzmann machine is a probabilistic graphic model that can be interpreted as a stochastic neural network. In practice, a Boltzmann machine is a model (including a certain number of parameters) that, when applied to a data distribution, is able to provide a representation. This model can be used to extract important aspects of an unknown distribution (target distribution) starting only from a sample of the latter.
Finally, an autoencoder was used for handwritten digit recognition. An autoencoder is a neural network whose purpose is to code its input into small dimensions, and the result obtained, to be able to reconstruct the input...