COMBINATORIAL OPTIMIZATION WITH FEEDBACK ARTIFICIAL NEURAL NETWORKS
Carsten Peterson
Abstract: A brief review is given for using feedback artificial neural
networks (ANN) to obtain good approximate solutions to combinatorial
optimization problems. The key element is the mean field approximation
(MFT) The methodology, which is illustrated for the graph bisection and
knapsack problems, is easily generalized to Potts systems. The latter is
related to the deformable templates method, which is illustrated with
the track finding problem. MFT is based on a variational principle, which
also can be generalized to non-integer problems.
Proceedings of ICANN '95 International Conference on Artificial
Neural Networks, October 1995, Paris, France , eds. F. Fogelman-Soulie and
P. Gallinari, EC2 \& Cie (Paris 1995).