Topographie Map Formation as Statistieal Inferenee
Roland Baddeley · 1995
Neurons representing similar aspects of the world are often found elose together in the cortex. It is proposed that this phenomenon can be modelIed using a statistical approach. We start by using a neural network to find the that were most likely to have generated the observed probability distribution of inputs. These features can be found using a Boltzmann machine architecture, but the results of this simple network are unsatisfactory. By adding two additional constraints (priors), that all representational units have the same probability of being true, and that nearby representational units are correlated, the network is shown to be capable of extracting distributed, spatially localised topographie representations based on an input of natural images. This is believed to be the first network capable of achieving this. As a model of topographie map formation, this framework has a num ber of strengths: 1) The framework is a general one, in which winner takes-all and distributed representations are special cases. 2) Though slow, the learning is simple and approximately Hebbian. 3) The network can extract topographie representations based on distributed input such as natural images.