Please use this identifier to cite or link to this item: http://hdl.handle.net/10553/73113
Title: The chemical metaphor in neural computation
Authors: Da Fonseca, J. Barahona
Da Fonseca, I. Barahona
Suárez Araujo, C. P. 
Da Fonseca, J. Simões
UNESCO Clasification: 2490 Neurociencias
120304 Inteligencia artificial
Issue Date: 2001
Journal: Lecture Notes in Computer Science 
Conference: 6th International Work-Conference on Artificial and Natural Neural Networks, IWANN 2001 
Abstract: Chemical reactions provide new insights concerning information processing in neural networks. Batch reactions are considered in first place and it is presented a model in which computational instructions will be conceptualised as being sent in parallel to multiple operational structures in sites where they are transformed in a way that generates a variety of computational viewpoints associated to a set of data. Anastomotic computations are considered as pre-processing operations implemented by convolutional weighing functions. Quantum valence theory is used to construct computational models for cognitive processes. The model of isolobal fragments is interpreted in a way that gives rise to classes of equivalence that represent concepts in extenso. The conscious quality of cognition is derived from the emergence of new macroscopic qualities from the relationships between valence bonds. Binding of representations is considered as dependent on dendro-dendritic relationships homeomorphic with orbital valence frontiers.
URI: http://hdl.handle.net/10553/73113
ISBN: 978-354042235-8
ISSN: 0302-9743
DOI: 10.1007/3-540-45720-8_22
Source: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) [ISSN 0302 -9743], v. 2084 LNCS (PART 1), p. 184-195, (Enero 2001)
Appears in Collections:Actas de congresos
Show full item record

Page view(s)

90
checked on Mar 3, 2024

Google ScholarTM

Check

Altmetric


Share



Export metadata



Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.