A modified probabilistic neural network signal processor for nonlinear signals

Author

Zaknich, Anthony ; Attikiouzel, Yianni

Author_Institution

Dept. of Electr. & Electron. Eng., Western Australia Univ., Nedlands, WA, Australia

Volume

1

fYear

1997

fDate

2-4 Jul 1997

Firstpage

291

Abstract

This paper introduces a practical and very effective network for nonlinear signal processing called the modified probabilistic neural network. It is a regression technique which uses a single radial basis function kernel whose bandwidth is related to the noise statistics. It has special advantages in application to time and spatial series signal processing problems because it is constructed directly and simply for the training signal waveform features. A sonar signal processing problem is used to illustrate its operation and to compare it with some other filters and neural networks

Keywords

Bayes methods; feedforward neural nets; learning (artificial intelligence); noise; parameter estimation; probability; sonar signal processing; statistical analysis; time series; Bayesian estimation theory; bandwidth; filters; modified probabilistic neural network signal processor; noise statistics; nonlinear signals; radial basis function kernel; regression technique; sonar signal processing; spatial series signal processing; time series signal processing; training signal waveform; Acoustic noise; Acoustical engineering; Bayesian methods; Equations; Estimation theory; Filters; Information processing; Intelligent systems; Neural networks; Signal processing;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Signal Processing Proceedings, 1997. DSP 97., 1997 13th International Conference on

Conference_Location

Santorini

Print_ISBN

0-7803-4137-6

Type

conf

DOI

10.1109/ICDSP.1997.628068

Filename

628068