Arbitrary Elastic Topologies and Ocular Dominance
Peter Dayan
Neural Computation, 5 392-401.
Abstract
The elastic net, which has been used to produce accounts of the
formation of topology preserving maps and ocular dominance columns
(OD), embodies a nearest neighbour topology. A Hebbian account of OD
is not so restricted -- and indeed makes the prediction that the width
of the columns depends on the nature of the (more general)
neighbourhood relations.
Elastic and Hebbian accounts have recently been unified -- raising a
question mark about their different determiners of column widths. This
paper considers this issue, and demonstrates theoretically that it is
possible to use more general topologies in the elastic net, including
those effectively adopted in the Hebbian model.
compressed postscript   pdf