Associative Memory Network Construction Algorithms


Bridgett, N.A., Brown, M., Mills, D.J. and Harris, C.J. (1994) Associative Memory Network Construction Algorithms. Int. Symp. on Signal Processing, Robotics And Neural Networks

Download

Full text not available from this repository.

Description/Abstract

In this paper a variety of Associative Memory Network (AMN) construction algorithms are briefly described and explained, namely the Multivariate Adaptive Regression Splines (MARS) algorithm of Friedman, and the Adaptive Spline Modelling of Observation Data (ASMOD) and Adaptive B-spline Basis function Modelling of Observation Data (ABBMOD) algorithms of Kavli. Such algorithms may be used for high-dimensional functional approximation in problems where there exits redundancy in the input data such that an accurate approximation may be formed additively from low order models. Since the process of model formation is automated, the algorithms may also be used for automatic model initialisation in some adaptive identification schemes. Since both the ASMOD and ABBMOD algorithms use B-spline basis functions which may be interpreted as a set of linguistic rules via fuzzy calculus, these algorithms may be used in the automatic generation of fuzzy rule bases.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Additional Information: Organisation: IMACS Address: Lille, France
Divisions: Faculty of Physical Sciences and Engineering > Electronics and Computer Science > Comms, Signal Processing & Control
ePrint ID: 250236
Date Deposited: 04 May 1999
Last Modified: 27 Mar 2014 19:51
Further Information:Google Scholar
URI: http://eprints.soton.ac.uk/id/eprint/250236

Actions (login required)

View Item View Item