Advertisement

Biological Cybernetics

, Volume 65, Issue 2, pp 129–133 | Cite as

Suboptimum solutions obtained by the Hopfield-Tank neural network algorithm

Article

Abstract

The neural network method of Hopfield and Tank claims to be able to find nearly-optimum solutions for discrete optimization problems, e.g. the travelling salesman problem. In the present paper, an example is given which shows that the Hopfield-Tank algorithm systematically prefers certain solutions even if the energy values of these solutions are clearly higher than the energy of the global minimum.

Keywords

Neural Network Global Minimum Travel Salesman Problem Discrete Optimization Suboptimum Solution 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Aarts EHL, Korst JHM (1987) Boltzmann machines and their applications. In: de Bakker et al. (eds) Parallel architectures and languages in Europe. Lecture Notes in Computer Science, vol 258. Springer, Berlin Heidelberg New York, pp 34–52Google Scholar
  2. Aarts EHL, Korst JHM (1989) Simulated annealing and Boltzmann machines. Wiley, Chichester New York BrisbaneGoogle Scholar
  3. Angéniol B, de La Croix Vaubois G, le Texier J-Y (1988) Self-organizing feature maps and the Travelling Salesman Problem. Neural Networks 1:289–293Google Scholar
  4. Cuykendall R, Reese R (1989) Scaling the neural network TSP algorithm. Biol Cybern 60:365–371Google Scholar
  5. Duque Antón M, Kunz D (1990) Parallel algorithms for channel assignment in cellular mobile radio systems: The neural network approach. In: Eckmiller R et al. (eds) Parallel processing in neural systems and computers. North-Holland, Amsterdam New York Oxford, pp 441–444Google Scholar
  6. Durbin R, Willshaw D (1987) An analogue approach to the travelling salesman problem using an elastic net method. Nature 326:689–691Google Scholar
  7. Guez A, Kam M, Eilbert JL (1989) Computational-complexity reduction for neural network algorithms. IEEE Trans Syst Man Cybern SMC-19:409–414Google Scholar
  8. Hopfield JJ (1982) Neural networks and physical systems with emergent collective computational abilities. Proc Natl Acad Sci USA 79:2554–2558Google Scholar
  9. Hopfield JJ (1984) Neurons with graded response have collective computational properties like those of two-state neurons. Proc Natl Acad Sci USA 81:3088–3092Google Scholar
  10. Hopfield JJ, Tank DW (1985) “Neural” computational of decisions in optimization problems. Biol Cybern 52:141–152Google Scholar
  11. Kirkpatrick S, Gelatt CD Jr, Vecchi MP (1983) Optimization by simulated annealing. Science 220:671–680Google Scholar
  12. Kohonen T (1984) Self-organization and associative memory. Springer Series in Information Sciences, vol 8. Springer, Berlin Heidelberg New YorkGoogle Scholar
  13. Kunz D (1990a) Channel assignment for cellular radio using neural networks. IEEE Trans Veh Techn: (to be published)Google Scholar
  14. Kunz D (1990b) Practical channel assignment using neural networks. 40th IEEE Veh Techn Conf, May 6–9, 1990, Orlando, FL, Proc. pp 652–655Google Scholar
  15. Lin S, Kernighan BW (1973) An effective heuristic algorithm for the travelling-salesman problem. Oper Res 21:498–516Google Scholar
  16. Ritter H, Schulten (1986) On the stationary state of Kohonen's self-organizing sensory mapping. Biol Cybern 54:99–106Google Scholar
  17. Van den bout DE, Miller TK III (1989) Improving the performance of the Hopfield-Tank neural network through normalization and annealing. Biol Cybern 62:129–139Google Scholar
  18. Wilson GV, Pawley GS (1988) On the stability of the Traveling Salesman algorithm of Hopfield and Tank. Biol Cybern 58:6370Google Scholar

Copyright information

© Springer-Verlag 1991

Authors and Affiliations

  • D. Kunz
    • 1
  1. 1.Philips GmbH Forschungslaboratorium AachenAachenGermany

Personalised recommendations