DBNs are graphical models which learn how to extract a deep hierarchical representation in the training knowledge. They model the joint distribution in between noticed vector Then, through the processes of gradient descent and backpropagation, the deep learning algorithm adjusts and matches a