Convergence properties of Kohonen's topology conserving maps: Fluctuations, stability, and dimension selection

DOI10.1007/BF00205972MaRDI QIDQ1108237zbMATH OpenFDO

Authors H. Ritter, Klaus Schulten

Publication date 1988

Published in Biological Cybernetics (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/BF00205972

zbMATH Keywords

automatic selection of feature dimensionsequilibrium mapFokker-Planck equationformal neuronsformation of topologically correct features mapsMarkovian algorithmpattern recognitionsignal processingstability

Mathematics Subject Classification ID

Applications of Brownian motions and diffusion theory (population genetics, absorption problems, etc.) (60J70) Learning and adaptive systems in artificial intelligence (68T05) Physiological, cellular and medical topics (92Cxx) Other natural sciences (mathematical treatment) (92F05)

Review text

We analyse a Markovian algorithm for the formation of topologically correct features maps proposed earlier by \textit{T. Kohonen} [ibid. 43, 59- 69 (1982; Zbl 0466.92002), and ``Selforganization and associative memory. (1984; Zbl 0528.68062)]. The maps from a space of input signals onto an array of formal neurons are generated by a learning scheme driven by a random sequence of input samples. The learning is described by an equivalent Fokker-Planck equation. Convergence to an equilibrium map can be ensured by a criterion for the time dependence of the learning step size. We investigate the stability of the equilibrium map and calculate the fluctuations around it. We also study an instability responsible for a phenomenon termed by Kohonen ``automatic selection of feature dimensions.

Recommendations

• Instabilities in Kohonen's self-organizing feature map
• scientific article; zbMATH DE number 447040
• On the rate of convergence in topology preserving neural networks
• Time behavior of topological ordering in self-organizing feature mapping
• On the topology distortion in self-organizing feature maps

Cited in

(24)

• Optimal interpolation of gravity maps using a modified neural network
• Self-organizing maps: Ordering, convergence properties and energy functions
• Heuristic and optimization approaches to extending the Kohonen self organizing algorithm
• On the topology distortion in self-organizing feature maps
• Formation of field discontinuities and islands in visual cortical maps
• Marginal median SOM for document organization and retrieval
• Study of the Kohonen network with a discrete state space
• A hybrid neural approach to combinatorial optimization
• Smoothed self-organizing map for robust clustering
• Convergence rates for sequences of bifurcation parameters of nonautonomous dynamical systems generated by flat top tent maps
• Self-organizing maps for imprecise data
• Multiple outlier detection in multivariate data using self-organizing maps
• SOM's mathematics
• Transition times in self-organizing maps
• Linear Geometric ICA: Fundamentals and Algorithms
• scientific article; zbMATH DE number 1555927 (Why is no real title available?)
• An adaptive neuromorphic model of ocular dominance map using floating gate `synapse'
• Winner-Relaxing Self-Organizing Maps
• Instabilities in Kohonen's self-organizing feature map
• Fault detection of systems with redundant sensors using constrained Kohonen networks
• Convergence of Map Seeking Circuits
• Convergence theorems for the Kohonen feature mapping algorithms with VLRPs
• On the rate of convergence in topology preserving neural networks
• An analysis of Kohonen's self-organizing maps using a system of energy functions

This page was built for publication: Convergence properties of Kohonen's topology conserving maps: Fluctuations, stability, and dimension selection

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1108237)