On hybrid connectionist-symbolic models

There are many similarities between grammar systems and artificial neural networks: parallelism, independently working elements (grammars/neurons), communication of the elements, absence of centralized control. On the other hand, there are crucial differences between symbolic and quantitative data p...

Full description

Saved in:
Bibliographic Details
Main Author: Sosík Petr
Format: Article
Published: 1997
Series:Acta cybernetica 13 No. 2
Kulcsszavak:Számítástechnika, Kibernetika
Subjects:
Online Access:http://acta.bibl.u-szeged.hu/12584
Description
Summary:There are many similarities between grammar systems and artificial neural networks: parallelism, independently working elements (grammars/neurons), communication of the elements, absence of centralized control. On the other hand, there are crucial differences between symbolic and quantitative data processing. We will try to give a brief overview of the methods of "building bridges" between symbolic and connectionist paradigm and to propose some recent results. After that, we will touch some essential problems, concerning incorporation of accepting/generating grammar systems and neural network models.
Physical Description:159-172
ISSN:0324-721X