Circular Backpropagation Networks for Classification
Sandro Ridella, Stefano Rovetta, and Rodolfo Zunino


IEEE Transactions on Neural Networks, vol.8, no.1, January 1997
(c) 1997 IEEE

ABSTRACT

The class of mapping networks is a general family of tools to perform
a wide variety of tasks; however, no unifying framework exists to describe
their theoretical and practical properties. This paper presents a
standardized, uniform representation for this class of networks, and
introduces a simple modification of the multilayer perceptron with
interesting properties, especially well suited to cope with pattern
classification tasks. The proposed model unifies the two main
representation paradigms found in the class of mapping networks for
classification, namely, the surface-based and the prototype-based schemes,
while retaining the advantage of being trainable by backpropagation.
The enhancement in the representation properties and the generalization
performance are assessed through results about the worst-case
requirement in terms of hidden units and about the Vapnik-Chervonenkis
dimension and Cover capacity. The theoretical properties of the network
also suggest that the proposed modification to the multilayer perceptron
is in many senses optimal. A number of experimental verifications also
confirm theoretical results about the model's inceased performances, as
compared with the multilayer perceptronand the Gaussian radial basis
function network.


INDEX TERMS

Feedforward neural networks, backpropagation, pattern classification,
knowledge representation