SURVEY OF DYNAMIC NEURAL NETWORK TECHNIQUES WITH APPLICATION TO TEMPORAL PROCESSING TASKS

PDF

Authors
  1. Stevenson, M.
Corporate Authors
Defence Research Establishment Atlantic, Dartmouth NS (CAN);New Brunswick Univ, Fredericton NB (CAN) Center for Research in Engineering and Applied Science
Abstract
Dynamic neural networks and their applications to temporal processing tasks are reviewed. Of special interest is a simple dynamic feedforward network known as the Finite Impulse Response Neural Network (FIRNN) and the classification of temporal patterns such as acoustic transients. Basic principles of the FIRNN are presented along with an explanation of how the FIRNN processes temporal information. Other dynamic networks which are discussed include additional dynamic feedforward neural networks, locally-recurrent globally-feedforward neural networks, partially-recurrent neural networks, and fully-recurrent neural networks. Application areas to which the various neural network paradigms have been applied include the classification of acoustic transients, prediction of chaotic time series, prediction and classification of speech signals, and modeling of nonlinear autoregressive (NAR) and nonlinear autoregressive moving-average (NAR-MA) processes.
Keywords
Finite impulse response;Dynamic neural networks;Transient signals;Sonar Information Management;Finite-duration Impulse Response (FIR)
Report Number
DREA-CR-97-458 — Contractor Report
Date of publication
01 Apr 1995
Number of Pages
75
DSTKIM No
98-01944
CANDIS No
508787
Format(s):
Hardcopy;Document Image stored on Optical Disk

Permanent link

Document 1 of 1

Date modified: