A simple algorithm that discovers efficient perceptual codes

We describe the 'wake-sleep' algorithm that allows a multilayer, unsupervised, neural network to build a hierarchy of representations of sensory input. The network has bottom-up 'recognition' connections that are used to convert sensory input into underlying representations. Unlike most artificial neural networks, it also has top-down 'generative' connections that can be used to reconstruct the sensory input from the representations. In the 'wake' phase of the learning algorithm, the network is driven by the bottom-up recognition connections and the top-down generative connections are trained to be better at reconstructing the sensory input from the representation chosen by the recognition process. In the 'sleep' phase, the network is driven top-down by the generative connections to produce a fantasized representation and a fantasized sensory input. The recognition connections are then trained to be better at recovering the fantasized representation from the fantasized sensory input. In both phases, the synaptic learning rule is simple and local. The combined effect of the two phases is to create representations of the sensory input that are efficient in the following sense: On average, it takes more bits to describe each sensory input vector directly than to first describe the representation of the sensory input chosen by the recognition process and then describe the difference between the sensory input and its reconstruction from the chosen representation