Accession Number : ADA433732


Title :   Lie Group Techniques for Neural Learning


Descriptive Note : Briefing charts


Corporate Author : SINTEF APPLIED MATHEMATICS OSLO (NORWAY)


Personal Author(s) : Celledoni, Elena


Full Text : http://www.dtic.mil/get-tr-doc/pdf?AD=ADA433732


Report Date : 03 JAN 2005


Pagination or Media Count : 37


Abstract : OUTLINE: 1. Neural Networks: a short introduction; 2. Independent Component Analysis: (a) Stochastic signal processing; (b) Constraint optimization in ICA; 3. Geometric Integration of Learning Equations: (a) Gradient flows and algorithms on manifolds; (b) MEC learning; (c) Newton methods; (d) Diffusion algorithms.


Descriptors :   *NEURAL NETS , *REASONING , *LEARNING MACHINES , *PATTERN RECOGNITION , *LIE GROUPS , *NEUROBIOLOGY , ALGORITHMS , SIGNAL PROCESSING , OPTIMIZATION , STOCHASTIC PROCESSES , INTEGRATION , DIFFERENTIAL EQUATIONS , DIFFUSION , INTEGRAL EQUATIONS , WORKSHOPS , FACTOR ANALYSIS , MANIFOLDS(MATHEMATICS).


Subject Categories : THEORETICAL MATHEMATICS
      CYBERNETICS


Distribution Statement : APPROVED FOR PUBLIC RELEASE