Neural Processing Letters

, Volume 39, Issue 1, pp 1–24 | Cite as

Self-Organizing Map Formation with a Selectively Refractory Neighborhood

Article

Abstract

Decreasing neighborhood with distance has been identified as one of a few conditions to achieve final states in the self-organizing map (SOM) that resemble the distribution of high-dimensional input data. In the classic SOM model, best matching units (BMU) decrease their influence area as a function of distance. We introduce a modification to the SOM algorithm in which neighborhood is contemplated from the point of view of affected units, not from the view of BMUs. In our proposal, neighborhood for BMUs is not reduced, instead the rest of the units exclude some BMUs from affecting them. Each neuron identifies, from the set of BMUs that influenced it in previous epochs, those to whom it becomes refractory to for the rest of the process. Despite that the condition of decreasing neighborhood over distance is not maintained, self-organization still persists, as shown by several experiments. The maps achieved by the proposed modification have, in many cases, a lower error measure than the maps formed by SOM. Also, the model is able to remove discontinuities (kinks) from the map in a very small number of epochs, which contrasts with the original SOM model.

Keywords

Self-organizing maps Asymmetric neighborhood  Map formation 

References

  1. 1.
    Cottrell M, Fort JC, Pagés G (1998) Theoretical aspects of the SOM algorithm. Neurocomputing 21:119–138CrossRefMATHGoogle Scholar
  2. 2.
    Kohonen T (2000) Self-Organizing maps, 3rd edn. Springer, New YorkGoogle Scholar
  3. 3.
    Ritter H (1999) Self-organizing maps on non-euclidean spaces. In: Oja E, Kaski S (eds) Kohonen maps. Elsevier, Amsterdam, pp 97–108Google Scholar
  4. 4.
    Yin Hujun (2008) The self-organizing maps: background, theories, extensions and applications. Computational intelligence: a compendium 2008:715–762Google Scholar
  5. 5.
    Van Hulle M (1997) Topology-preserving map formation achieved with a purely local unsupervised competitive learning rule. Neural Netw 10(3):431–446CrossRefGoogle Scholar
  6. 6.
    Flanagan J (1999) Sufficient conditions for self-organization in the SOM with a decreasing neighborhood function of any width. In: Conference of artificial neural networks. Conference Pub 470Google Scholar
  7. 7.
    Erwin E, Obermayer K, Schulten K (1992) Self-organizing maps: ordering. convergence properties and energy functions. Biol Cyb 67:47–55CrossRefMATHGoogle Scholar
  8. 8.
    Berglund E, Sitte J (2006) The parameterless self-organizing map algorithm. IEEE Trans Neural Netw 17(2):305–316CrossRefGoogle Scholar
  9. 9.
    Erwin E, Obermayer K, Schulten K (1992b) self-organizing maps: stationary states, metastability and convergence rate. Biol Cyb 67:35–45CrossRefMATHGoogle Scholar
  10. 10.
    Bamford SA, Af Murray (2010) Synaptic rewiring for topographic map formation and receptive field development. Neural Netw 23:517–527CrossRefGoogle Scholar
  11. 11.
    Maia JEB, Barreto GA, Coelho ALV (2011) Visual object tracking by an evolutionary self-organizing neural network. J Intell Fuzzy Syst 22(2–3):69–81MathSciNetGoogle Scholar
  12. 12.
    Willshaw D, Malsburg C (1976) How patterned neural connections can be set up by self-organization. Proc R Soc Lond 194:431–445Google Scholar
  13. 13.
    Buzsaki G (2010) Neural syntax: cell assemblies, synapsembles, and readers. Neuron 68(3):362–85.Google Scholar
  14. 14.
    Miikkulainen R, Bednar J, Choe, S (2005) Computational maps in the visual cortex. Springer, New YorkGoogle Scholar
  15. 15.
    Forciniti L, Schmidt CE, Zaman H (2009) Computational model provides insight into the distinct responses of neurons to chemical and topographical cues. Ann Biomed Eng 37(2):363–374Google Scholar
  16. 16.
    Tamarit F, Stariolo D, Cannas A, Serras P (1996) Effects of refractory periods in the dynamics of a diluted neural network. Phys Rev E 53:51465152CrossRefGoogle Scholar
  17. 17.
    Neme A, Hernández S, Neme O (2011) An electoral preferences model based on self-organizing maps. J Comput Sci 2(4):345–352CrossRefGoogle Scholar
  18. 18.
    Neme A, Hernández S, Neme O, Hernández (2009) Self-organizing maps with non-cooperative strategies. In: Advances in self-organizing maps. Springer, New York, pp 200–208Google Scholar
  19. 19.
    Trappenberg T, Hartono P, Rasmusson D (2009) Top-down control of learning in biological self-organizing maps. WSOM 2009:316–324Google Scholar
  20. 20.
    Aoki T, Aoyagi T (2007) Self-organizing maps with asymmetric neighborhood function. Neural Comput 19(9):2515–2535CrossRefMATHMathSciNetGoogle Scholar
  21. 21.
    Neme A, Miramontes P (2006) A parameter in the SOM learning rule that incorporates activation frequency. ICANN 1:455–463Google Scholar
  22. 22.
    Lee J, Verleysen M (2002) Self-organizing maps with recursive neighborhood adaption. Neural Netw 15:993–1003CrossRefGoogle Scholar
  23. 23.
    Bauer H, Pawelzik K (1992) Quantifying the neighborhood preservation in self-organizing feature maps. IEEE Trans Neural Netw 3(4):570–579CrossRefGoogle Scholar
  24. 24.
    Bauer H, Herrmann M, Villmann T (1999) Neural maps and topographic vector quantization. Neural Netw 12(4–5):659–676CrossRefGoogle Scholar
  25. 25.
    Kiviluoto K (1996) Topology preservation in self-organizing maps. In: Proceedings of ICNN96, IEEE international conference on neural networksGoogle Scholar
  26. 26.
    Venna J, Kaski S (2001) Neighborhood preservation in nonlinear projection methods: an experimental study. In: Dorffner G, Bischof H, Hornik K (eds) ICANN 2001, vol 2130. Springer, New York, pp 485-491Google Scholar
  27. 27.
    Mayer R, Neumayer R, Baum D, Rauber A (2009) Analytic comparison of self-organising maps. In: WSOM 2009. LNCS, vol 5629, pp 182–190Google Scholar
  28. 28.
    Cover T Thomas J (2006) Elements of information theory. Wiley, New YorkGoogle Scholar
  29. 29.
    Cellucci CJ, Albano AM, Rapp PE (2005) Statistical validation of mutual information calculations: comparison of alternative numerical algorithms. Phys Rev E 71:066208CrossRefGoogle Scholar
  30. 30.
    Neme A, Nido A, Mireles V, Miramontes P (2008) The self-organized chaos game representation for genomic signatures analysis. Learn Nonlinear Models 6(2):111–120Google Scholar
  31. 31.
    Almeida J, Carrico J, Maretzek A, Noble P, Fletcher M (2001) Analysis of genomic sequences by chaos game representatio. Bioinformatics 17(5):429–437CrossRefGoogle Scholar
  32. 32.
    Jeffrey J (1998) Chaos game representation of sequences in Chaos and Fractals: a computer graphical journey. PickoverGoogle Scholar
  33. 33.
    Carreón G, Hernández E, Miramontes P (2005) DNA circular game of chaos. In: Uribe F, Garcia-Colín L (eds) Statistical physics and beyond, American Institute of PhysicsGoogle Scholar
  34. 34.
    Aoki T, Ota K, Kurata K (2009) Ordering process of self-organizing maps improved by asymmetric neighborhood function. Cogn Neurodyn 3:9–15. doi:10.1007/s11571-008-9060-2 CrossRefGoogle Scholar
  35. 35.
    Murakochi K, Sato Y (2007) Reducing topological defects in self-organizing maps using multiple scale neighborhood functions. Biosystems 90–1:101–104CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  1. 1.Complex Systems GroupUniversidad Autónoma de la Ciudad de MéxicoMexicoMexico
  2. 2.Institute for Molecular Medicine Finland (FIMM)HelsinkiFinland
  3. 3.Faculty of SciencesUniversidad Nacional Autónoma de MéxicoMéxico CityMexico
  4. 4.Centre for Complexity SciencesUniversidad Nacional Autónoma de MéxicoMéxico CityMexico

Personalised recommendations