Number of the records: 1  

Innovations in Neural Information Paradigms and Applications

  1. 1.
    0328492 - ÚI 2010 RIV DE eng M - Monography Chapter
    Kůrková, Věra
    Estimates of Model Complexity in Neural-Network Learning.
    [Odhady modelové složitosti při učení neuronových sítí.]
    Innovations in Neural Information Paradigms and Applications. Berlin: Springer, 2009 - (Bianchini, M.; Maggini, M.; Scarselli, F.; Jain, L.), s. 97-111. Studies in Computational Intelligence, 247. ISBN 978-3-642-04002-3
    R&D Projects: GA MŠMT(CZ) 1M0567
    Institutional research plan: CEZ:AV0Z10300504
    Keywords : model complexity * neural networks * learning from data
    Subject RIV: IN - Informatics, Computer Science

    Model complexity in neural-network learning is investigated using tools from nonlinear approximation and integration theory. Estimates of network complexity are obtained from inspection of upper bounds on convergence of minima of error functionals over networks with an increasing number of units to their global minima. The estimates are derived using integral transforms induced by computational units. The role of dimensionality of training data defining error functionals is discussed.

    Modelová složitost při učení neuronových sítí je studována pomocí metod z teorie nelineární aproximace a integrace. Odhady složitosti sítí jsou odvozeny z horních odhadů rychlosti konvergence minim chybových funkcionálů dosažitelných na sítích s rostoucím počtem výpočetních jednotek.
    Permanent Link: http://hdl.handle.net/11104/0174795

     
     
Number of the records: 1  

  This site uses cookies to make them easier to browse. Learn more about how we use cookies.