HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation

Wavelets in identification wavelets, splines, neurons, fuzzies : how good for identification

Anatoli Juditsky 1 Qinghua Zhang 1 Bernard Delyon 1 Pierre-Yves Glorennec 1 Albert Benveniste 1
1 AS - Signal Processing and Control
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, INRIA Rennes
Abstract : This is a tutorial about nonparametric nonlinear system identification. Advantages and limitations of this approach are discussed from the engineer's point of view. Classical as well as modern techniques are discussed, this includes kernel and projection estimates, neural networks and hinging hyperplanes, and mainly wavelet estimators. Both practical and mathematical issues are investigated. Advantages and limitations of wavelet based techniques are emphazised. Finally we show how fuzzy models may play a role in this game, as a framework for expressing prior knowledge on the system. The whole material is illustrated on some application examples.
Document type :
Complete list of metadata

Contributor : Rapport de Recherche Inria Connect in order to contact the contributor
Submitted on : Wednesday, May 24, 2006 - 3:08:51 PM
Last modification on : Friday, February 4, 2022 - 3:30:57 AM
Long-term archiving on: : Sunday, April 4, 2010 - 9:28:23 PM


  • HAL Id : inria-00074359, version 1


Anatoli Juditsky, Qinghua Zhang, Bernard Delyon, Pierre-Yves Glorennec, Albert Benveniste. Wavelets in identification wavelets, splines, neurons, fuzzies : how good for identification. [Research Report] RR-2315, INRIA. 1994. ⟨inria-00074359⟩



Record views


Files downloads