Abstract
Guided by the neurobiological principles of self-organization and population coding, we develop a simple, neural, one-layer model for auto-association. Its core is a feature map endowed with self-organized lateral connections. Input patterns are coded by small spots of active neurons. The time evolution of neural activity then realizes an auto-association process by a recurrent attractor dynamics. Population coding is preserved due to a balance of diffusive spreading of activity and competitive refocusing. Because of its simplicity, the model allows a thorough qualitative and quantitative understanding. We show that the network is capable of performing a cluster analysis and hierarchical classification of data and, thus, qualifies as a tool for unsupervised statistical data analysis.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Abramowitz M, Stegun IA (eds) (1984) Handbook of mathematical functions. Harry Deutsch Verlag, Thun
Adolphs U (1992) Selbstorganisierende Perzeptronstrukturen mit radialen Basisfunktionen. Diplomarbeit Physik-Department, Technische Universität München
Baldi P, Heiligenberg W (1988) How sensory maps could enhance resolution through ordered arrangements of broadly tune receivers. Biol Cybern 59:313–318
Cooper L (1974) A possible organization of animal memory and learning. In: Lundquist B, Lindquist S (eds) Proc Nobel Symp Collective Properties of Physical Systems. Acad Press, New York
Dersch DR, Tavan P (1994) Asymptotic level density in topological feature maps. IEEE Trans Neural Net, in press
Duda RO, Hart PE (1973) Pattern classification and scene analysis. Wiley, New York
Gardiner CW (1990) Handbook of stochastic methods. Springer, Berlin Heidelberg New York
Haken H (1983) Synergetics. An introduction. Springer, Berlin Heidelberg New York
Hemmen JL van, Joffe L, Kühn R, Vaas M (1990) Increasing the efficiency of a neural network through unlearning. Physica 163A: 386–392
Hopfield JJ (1982) Neural networks and physical systems with emergent collective computational abilities. Proc Natl Acad Sci USA 79:2554–2558
Hopfield JJ, Feinstein DI, Palmer RG (1983) ‘Unlearning’ has a stabilizing effect on collective memories. Nature 304:158–159
Kohonen T (1982) Self-organized formation of topologically correct feature maps. Biol Cybern 43:59–69
Kosko B (1988) Bidirectional associative memories. IEEE Trans Syst Man Cybern 18:49–60
Kunstmann N, Hillermeier C, Rabus B, Tavan P (1994) An associative memory that can form hypotheses—a phase-coded neural network. Biol Cybern, in press
Kunstmann N, Hillermeier C, Tavan P (1993) Associative memories that can form hypotheses: phased-coded network architectures. Proc ICANN, Amsterdam
Linde Y, Buzo A, Gray RM (1980) An algorithm for vector quantizer design. IEEE Trans Comm 28:84–95
Malsburg C von der (1981) The correlation theory of brain function. (Technical Report 81–2) Department für Neurobiologie, MaxPlanck-Institut für Biophysikalische Chemie, Göttingen
Moody J, Darken C (1989) Fast learning in networks of locally-tuned processing units. Neural Comput 1:281–294
Poggio T, Girosi F (1990) Regularization algorithms for learning that are equivalent to multilayer networks. Science 247: 978–982
Rabus B (1992) Hypothesenbildung in phasenkodierten Assoziativspeichern. Diplomarbeit Physik-Department, Technische Universität München
Reichl LE (1980) A modern course in statistical physics. University of Texas Press, Austin.
Ritter H (1991) Asymptotic level density for a class of vector quantization processes. IEEE Trans Neural Net 1:173–175
Sparks D, Lee C, Rohrer W (1990) Population coding of the direction, amplitude, and velocity of saccadic eye-movements in the superior colliculus. Cold Spring Harbor Symp Quant Biol 55:805–811
Tavan P, Grubmüller H, Kühnel H (1990) Self-organization of associative memory and pattern classification: recurrent signal processing on topological feature maps. Biol Cybern 64:95–105
Willshaw DJ, Malsburg C von der (1976) How patterned neural connections can be set up by self-organization. Proc R Soc Lond [Biol] 194:431–445
Zador PL (1982) Asymptotic quantization error of continuous signals and the quantization dimension. IEEE Trans Inform Theor 28:139–149
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Hillermeier, C., Kunstmann, N., Rabus, B. et al. Topological feature maps with self-organized lateral connections: a population-coded, one-layer model of associative memory. Biol. Cybern. 72, 103–117 (1994). https://doi.org/10.1007/BF00205975
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1007/BF00205975