Using Wavelet Network in Nonparametric Estimation Qinghua

In this paper one approach is proposed for using wavelets in non parametric regression estimation. The proposed non parametric estimator, named wavelet network, has a neural network like structure, but consists of wavelets. It makes use of techniques of regressor selection completed with backpropagation procedures. It is capable of handling nonlinear regressions of moderately large input dimension with sparse training data. Numerical examples are reported to illustrate the performance of this proposed approach. Utiliser le r eseau d'ondelettes dans estimation non param etrique R esum e : Dans ce document une approche est propos ee pour utiliser des onde-lettes en estimation non param etrique. L'estimateur propos e, appel e r eseau d'on-delettes, a une structure similaire a celle de certains r eseaux de neurones. Il utilise des techniques de la s election de regresseurs compl et ees par des proc edures dites \backpropagation". Il est capable de traiter des probl emes de regression non lin eaire de dimensions mod er ees avec des donn ees d'apprentissage creuses. Des exemples num eriques sont pr esent es pour illustrer la performance de l'approche propos ee.

[1]  R. Duffin,et al.  A class of nonharmonic Fourier series , 1952 .

[2]  H. Akaike Fitting autoregressive models for prediction , 1969 .

[3]  Shie Qian,et al.  Signal representation using adaptive normalized Gaussian functions , 1994, Signal Process..

[4]  Qinghua Zhang,et al.  Early Warning of Slight Changes in Systems and Plants , 1993 .

[5]  Qinghua Zhang Contribution à la surveillance de procédés industriels , 1991 .

[6]  L. Griffiths Introduction to discrete-time signal processing , 1978 .

[7]  J. Rissanen,et al.  Modeling By Shortest Data Description* , 1978, Autom..

[8]  Shang-Liang Chen,et al.  Orthogonal least squares learning algorithm for radial basis function networks , 1991, IEEE Trans. Neural Networks.

[9]  Andrew R. Barron,et al.  Universal approximation bounds for superpositions of a sigmoidal function , 1993, IEEE Trans. Inf. Theory.

[10]  S. Mallat Multiresolution approximations and wavelet orthonormal bases of L^2(R) , 1989 .

[11]  Louis A. Hageman,et al.  Iterative Solution of Large Linear Systems. , 1971 .

[12]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[13]  Qinghua Zhang,et al.  Regressor selection and wavelet network construction , 1993, Proceedings of 32nd IEEE Conference on Decision and Control.

[14]  F. Girosi,et al.  Networks for approximation and learning , 1990, Proc. IEEE.

[15]  C. Chui Wavelets: A Tutorial in Theory and Applications , 1992 .

[16]  Sheng Chen,et al.  Orthogonal least squares methods and their application to non-linear system identification , 1989 .

[17]  Ingrid Daubechies,et al.  Ten Lectures on Wavelets , 1992 .

[18]  Bhavik R. Bakshi,et al.  Wave‐net: a multiresolution, hierarchical neural network with localized learning , 1993 .

[19]  Stéphane Mallat,et al.  Matching pursuits with time-frequency dictionaries , 1993, IEEE Trans. Signal Process..

[20]  J. Friedman,et al.  Projection Pursuit Regression , 1981 .