Deep learning

Autoencoders: Sparse and Deep

A typical machine learning situation assumes we have a large number of training vectors, for example gray level images of 16×16 size representing digits 0 to 9 with each.