Document Detail

Neural net pruning based on functional behavior of neurons.
MedLine Citation:
PMID:  8044375     Owner:  NLM     Status:  MEDLINE    
This paper proposes a new pruning method based on merging neurons with similar functional behavior which is defined by the internal representations of each neuron for the entire training set. Classification of neurons by their functional behavior with respect to the input vectors provides a powerful tool for pruning neurons and connections, thus reducing the network complexity and increasing its generalization capability. The most remarkable property of this pruning scheme is its ability to preserve net functionality by transferring the role of every removed neuron to the most fitted neuron of the surviving ones, using a unique merging and compensation procedure. The implementation of the proposed method is demonstrated using a detailed numerical example and its performance is examined by a statistical measure calculated by repeating the training procedure several times. The influence of parameter selection on pruning performance and generalization ability is discussed and demonstrated by examining statistical results.
N Shamir; D Saad; E Marom
Related Documents :
7527075 - Modulation of associative memory function in a biophysical simulation of rat piriform c...
11918995 - Collective bursting in layer iv. synchronization by small thalamic inputs and recurrent...
23435215 - Dynamic causal modeling suggests serial processing of tactile vibratory stimuli in the ...
8894125 - A model for temporal and intensity coding in insect olfaction by a network of inhibitor...
11426165 - Load sharing within a human lumbar vertebral body using the finite element method.
23825935 - Dynamic excitatory and inhibitory gain modulation can produce flexible, robust and opti...
Publication Detail:
Type:  Journal Article    
Journal Detail:
Title:  International journal of neural systems     Volume:  4     ISSN:  0129-0657     ISO Abbreviation:  Int J Neural Syst     Publication Date:  1993 Jun 
Date Detail:
Created Date:  1994-08-31     Completed Date:  1994-08-31     Revised Date:  2004-11-17    
Medline Journal Info:
Nlm Unique ID:  9100527     Medline TA:  Int J Neural Syst     Country:  SINGAPORE    
Other Details:
Languages:  eng     Pagination:  143-58     Citation Subset:  IM    
Faculty of Engineering, Tel Aviv University, Ramat Aviv, Israel.
Export Citation:
APA/MLA Format     Download EndNote     Download BibTex
MeSH Terms
Neural Networks (Computer)*
Neurons / physiology*

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

Previous Document:  Control of the complexity of associative memory dynamics by neuronal adaptation.
Next Document:  A quantitative study of pruning by optimal brain damage.