Training a deep network (a quick overview)
	• First train a layer of features that receive input directly from the pixels. – This is done by training a restricted Boltzmann machine. • Then treat the activations of the trained features as if they were pixels and learn features of features in a second hidden layer. • Each time we add another layer of features we improve a variational bound on how well we are modeling the set of training images. – This assumes that the layers do not get smaller and they are initialized correctly.