January 2018
Beginner to intermediate
284 pages
8h 35m
English
A RBM is a neural network with only two layers: the visible layer and the hidden layer. Each visible node/neuron is connected to each hidden node. Restriction means there is no intra-layer communication, that is, there are no connections between the visible-visible nodes or the hidden-hidden nodes. It was one of the earliest models to be introduced in the area of artificial intelligence and has been applied successfully in many areas such as dimensionality reduction, classification, feature learning, and anomaly detection.
The following figure shows its basic structure:

Read now
Unlock full access