Response analysis of feed-forward neural network predictors

Author

Varone, Barbara ; Tanskanen, Jarono M A ; Ovaska, Seppo J.

Author_Institution

Politecnico di Milano, Italy

Volume

4

fYear

1997

fDate

21-24 Apr 1997

Firstpage

3309

Abstract

We investigate the characteristics of some one-step-ahead nonlinear predictors based on a two-layer feed-forward neural network (2LFNN). The behavior of neural networks (NN) is investigated in the frequency domain using two frequency response estimation techniques, and in the time domain, by analyzing the unit step and triangular pulse responses. Some of the estimated frequency responses of these NNs resemble those of corresponding linear polynomial predictors, revealing the nearly polynomial nature of the applied training signals. The similarity of the two frequency response estimates is an indication of good generalization properties

Keywords

code division multiple access; feedforward neural nets; frequency estimation; frequency response; frequency-domain analysis; land mobile radio; learning (artificial intelligence); multi-access systems; polynomials; signal processing; step response; time-domain analysis; CDMA; feed-forward neural network predictors; frequency domain; frequency response analysis; frequency response estimation; generalization properties; linear polynomial predictors; mobile communications system; one step ahead nonlinear predictors; time domain; training signals; triangular pulse response; two-layer feedforward neural network; unit step response; Electronic mail; Feedforward neural networks; Feedforward systems; Frequency estimation; Frequency response; Multiaccess communication; Neural networks; Neurons; Power control; Time domain analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1997. ICASSP-97., 1997 IEEE International Conference on

Conference_Location

Munich

ISSN

1520-6149

Print_ISBN

0-8186-7919-0

Type

conf

DOI

10.1109/ICASSP.1997.595501

Filename

595501