Boltzmann Machines (BM)

Similar to Hopfield Networks (HN), BMs are fully connected neural networks; but unlike HNs, some neurons are marked as input neurons and others remain “hidden”. BMs are Stochastic Networks with binary activation.

The input neurons become output neurons at the end of a full network update. It starts with random weights and learns through back-propagation or contrastive divergence.

The activation is controlled by a global temperature value, which if lowered lowers the energy of the cells. This lower energy causes their activation patterns to stabilize. The network reaches an equilibrium given the right temperature.