Training Wavelet Networks for Nonlinear Dynamic Input-Output Modeling

Yacine Oussar (1) , Isabelle Rivals (1) , Léon Personnaz (1) , Gerard Dreyfus (1)

1 Laboratoire d'Electronique

In the framework of nonlinear process modeling, we propose training algorithms for feedback wavelet networks used as nonlinear dynamic models. An original initialization procedure is presented, that takes the locality of the wavelet functions into account. Results obtained for the modeling of several processes are presented; a comparison with networks of neurons with sigmoidal functions is performed.

Keywords: Training; Wavelet networks; Nonlinear dynamic modeling; Neural networks; Feedback networks; Recurrent networks

Domains: Statistics [stat]; Machine Learning [stat.ML]; Computer Science [cs]; Neural and Evolutionary Computing [cs.NE]

Yacine Oussar, Isabelle Rivals, Léon Personnaz, Gerard Dreyfus. Training Wavelet Networks for Nonlinear Dynamic Input-Output Modeling. Neurocomputing, 1998, 20 (1-3), pp.173-188. ⟨hal-00797616⟩