Document Detail


A neural support vector machine.
MedLine Citation:
PMID:  20092978     Owner:  NLM     Status:  MEDLINE    
Abstract/OtherAbstract:
Support vector machines are state-of-the-art pattern recognition algorithms that are well founded in optimization and generalization theory but not obviously applicable to the brain. This paper presents Bio-SVM, a biologically feasible support vector machine. An unstable associative memory oscillates between support vectors and interacts with a feed-forward classification pathway. Kernel neurons blend support vectors and sensory input. Downstream temporal integration generates the classification. Instant learning of surprising events and off-line tuning of support vector weights trains the system. Emotion-based learning, forgetting trivia, sleep and brain oscillations are phenomena that agree with the Bio-SVM model. A mapping to the olfactory system is suggested.
Authors:
Magnus Jändel
Related Documents :
17211628 - Modelling memory functions with recurrent neural networks consisting of input compensat...
25285398 - Using the threat probability task to assess anxiety and fear during uncertain and certa...
18542278 - Low memory distributed reconstruction of large digital holograms.
23581318 - Enhancing multiphoton rates with quantum memories.
16528428 - Audiologic management of older adults with hearing loss and compromised cognitive/psych...
661258 - Learning visual correlates of fricative production by normal-hearing subjects: a prelim...
Publication Detail:
Type:  Journal Article; Research Support, Non-U.S. Gov't     Date:  2010-01-11
Journal Detail:
Title:  Neural networks : the official journal of the International Neural Network Society     Volume:  23     ISSN:  1879-2782     ISO Abbreviation:  Neural Netw     Publication Date:  2010 Jun 
Date Detail:
Created Date:  2010-04-26     Completed Date:  2010-07-22     Revised Date:  -    
Medline Journal Info:
Nlm Unique ID:  8805018     Medline TA:  Neural Netw     Country:  United States    
Other Details:
Languages:  eng     Pagination:  607-13     Citation Subset:  IM    
Copyright Information:
Copyright 2010 Elsevier Ltd. All rights reserved.
Affiliation:
Agora for Biosystems, Box 57 SE-193 22, Sigtuna, Sweden; Swedish Defence Research Agency, SE-164 90, Stockholm, Sweden. magnus@jaendel.se
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Descriptor/Qualifier:
Algorithms*
Animals
Learning / physiology
Memory / physiology
Neural Networks (Computer)*
Neurons / physiology
Nonlinear Dynamics
Olfactory Pathways / physiology
Olfactory Perception / physiology
Pattern Recognition, Automated
Periodicity
Smell / physiology
Time Factors

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


Previous Document:  Plasma metabolic fingerprinting of childhood obesity by GC/MS in conjunction with multivariate stati...
Next Document:  Contralateral hyperperfusion and ipsilateral hypoperfusion by ictal SPECT in patients with mesial te...