Convergence of the one-dimensional Kohonen algorithm

We show in a very general framework the a.s. convergence of the one-dimensional Kohonen algorithm–after self-organization–to its unique equilibrium when the learning rate decreases to 0 in a suitable way. The main requirement is a log-concavity assumption on the stimuli distribution which includes all the usual (truncated) probability distributions (uniform, exponential, gamma distribution with parameter ≥ 1, etc.). For the constant step algorithm, the weak convergence of the invariant distributions towards equilibrium as the step goes to 0 is established too. The main ingredients of the proof are the Poincaré-Hopf Theorem and a result of Hirsch on the convergence of cooperative dynamical systems.

[1]  M. Benaïm.  A Dynamical System Approach to Stochastic Approximations , 1996 .

[2]  F. R. Gantmakher.  The Theory of Matrices , 1984 .

[3]  M. Benaïm.  Recursive algorithms, urn processes and chaining number of chain recurrent sets , 1998, Ergodic Theory and Dynamical Systems.

[4]  Hal L. Smith,et al.  Monotone Dynamical Systems: An Introduction To The Theory Of Competitive And Cooperative Systems (Mathematical Surveys And Monographs) By Hal L. Smith , 1995 .

[5]  M. Hirsch.  Systems of Differential Equations that are Competitive or Cooperative II: Convergence Almost Everywhere , 1985 .

[6]  J. Fort,et al.  On the A.S. Convergence of the Kohonen Algorithm with a General Neighborhood Function , 1995 .

[7]  Gilles Pagès,et al.  About the Kohonen algorithm: strong or weak self-organization? , 1996, ESANN.

[8]  G. Pagès,et al.  Self-organization and a.s. convergence of the one-dimensional Kohonen algorithm with non-uniformly distributed stimuli , 1993 .

[9]  Ali A. Sadeghi,et al.  Asymptotic behavior of self-organizing maps with nonuniform stimuli distribution , 1998 .

[10]  Gilles Pagès,et al.  Asymptotic Behavior of a Markovian Stochastic Algorithm with Constant Step , 1999 .

[11]  Teuvo Kohonen,et al.  Self-Organization and Associative Memory , 1988 .

[12]  M. Hirsch.  Stability and convergence in strongly monotone dynamical systems. , 1988 .

[13]  Gilles Pagès,et al.  CONVERGENCE OF STOCHASTIC ALGORITHMS: FROM THE KUSHNER-CLARK THEOREM TO THE LYAPOUNOV FUNCTIONAL METHOD , 1996 .

[14]  M. Métivier,et al.  Théorèmes de convergence presque sure pour une classe d'algorithmes stochastiques à pas décroissant , 1987 .