Document Detail


Genetic algorithm applied to hierarchically coupled associative memories.
MedLine Citation:
PMID:  20020348     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
Inspired by the theory of neuronal group selection (TNGS), we have carried out an analysis of the capacity of convergence of a multi-level associative memory based on coupled generalized-brain-state-in-a-box (GBSB) networks through evolutionary computation. The TNGS establishes that a memory process can be described as being organized functionally in hierarchical levels where higher levels coordinate sets of functions of lower levels. According to this theory, the most basic units in the cortical area of the brain are called neuronal groups or first-level blocks of memories and the higher-level memories are formed through selective strengthening or weakening of the synapses amongst the neuronal groups. In order to analyse this effect, we propose that the higher levels should emerge through a learning mechanism as correlations of lower level memories. According to this proposal, this paper describes a method of acquiring the inter-group synapses based on a genetic algorithm. Thus the results show that genetic algorithms are feasible as they allow the emergence of complex behaviours which could be potentially excluded in other learning process.
Authors:
Rog?rio Martins Gomes; Ant?nio P?dua Braga; Henrique E Borges
Related Documents :
21106078 - Neural networks engaged in tactile object manipulation: patterns of expression among he...
12678588 - Temporal dynamics of the human response to symmetry.
23581318 - Enhancing multiphoton rates with quantum memories.
20085698 - Making cellular memories.
22212588 - Face identity recognition in autism spectrum disorders: a review of behavioral studies.
19435628 - Individual differences in sternberg's memory scanning task.
Publication Detail:
Type:  Journal Article; Research Support, Non-U.S. Gov't    
Journal Detail:
Title:  Advances in experimental medicine and biology     Volume:  657     ISSN:  0065-2598     ISO Abbreviation:  Adv. Exp. Med. Biol.     Publication Date:  2010  
Date Detail:
Created Date:  2009-12-18     Completed Date:  2010-04-08     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  0121103     Medline TA:  Adv Exp Med Biol     Country:  United States    
Other Details:
Languages:  eng     Pagination:  187-99     Citation Subset:  IM    
Affiliation:
CEFET-MG, Av. Amazonas 7675, Belo Horizonte, MG, CEP 30510-000, Brazil. rogerio@lsi.cefetmg.br
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Algorithms*
Association Learning / physiology*
Computer Simulation
Humans
Memory / physiology*
Models, Neurological*
Neural Networks (Computer)

From MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine


Previous Document:  Neuromorphic systems: past, present and future.
Next Document:  Vector quantization of speech frames based on self-organizing maps.