Learning of RBM (Contrastive Divergence (Hinton, 2002))

18th-October-2020

What if I quit 𝑇𝑇 at finite times? It corresponds to the minimization of the objective function called Contrastive Divergence.

Learning of RBM (Contrastive Divergence (Hinton, 2002))

The parameter updating rule in the case of terminating 𝑇𝑇 at finite number of times corresponds to the minimization with Contrastive Divergence as the objective function. • RBM can obtain accurate learning results even if 𝑇𝑇 = 1