DocumentCode
275946
Title
A framework of hierarchy for neural theory
Author
Vellacott, O.R.
Author_Institution
Edinburgh Univ., UK
fYear
1991
fDate
18-20 Nov 1991
Firstpage
237
Lastpage
241
Abstract
Describes a framework of hierarchy within which representations of neural functions and data may be transformed from level to level. Within this framework, the function of a neural system may be abstracted above the primitive level so that it may be understood at arbitrary higher conceptual levels. The framework is natural to neural systems in that representations are completely distributed at each level of abstraction. Neural Compilation, the process by which a hierarchical specification of a neural system is implemented, is facilitated by ANNECS, a software tool. What is significant about this framework is that it enables one to comprehend neural systems at arbitrary levels of abstraction. As discussed, this ability is essential to be able to understand the operation of any nontrivial system and should aid analysis of neural systems by raising representations above the primitive level
Keywords
neural nets; software tools; ANNECS; framework of hierarchy; neural functions; neural theory; software tool;
fLanguage
English
Publisher
iet
Conference_Titel
Artificial Neural Networks, 1991., Second International Conference on
Conference_Location
Bournemouth
Print_ISBN
0-85296-531-1
Type
conf
Filename
140323
Link To Document