Deterministic Crowding Helps to Evolve Non-correlated Active Neurons
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F05%3A03115266" target="_blank" >RIV/68407700:21230/05:03115266 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Deterministic Crowding Helps to Evolve Non-correlated Active Neurons
Original language description
In the original Multilayered Iterative Algorithm the exhaustive search is used to find and select units with the best transfer function, connected to most relevant inputs. Recently, several modifications using standard genetic algorithms instead of exhaustive search appeared. This paper shows how to improve the efficiency of search by evolving non-correlated units (active neurons). This is attained by employing Deterministic Crowding (DC) method proposed by Mahfoud in 1995. As a by-product of using DC method, we can estimate the importance of input variables for modelled output (feature ranking).
Czech name
Není k dispozici
Czech description
Není k dispozici
Classification
Type
D - Article in proceedings
CEP classification
JC - Computer hardware and software
OECD FORD branch
—
Result continuities
Project
—
Continuities
Z - Vyzkumny zamer (s odkazem do CEZ)
Others
Publication year
2005
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceedings of the International Workshop on Inductive Modeling IWIM-2005
ISBN
966-02-3734-0
ISSN
—
e-ISSN
—
Number of pages
8
Pages from-to
21-28
Publisher name
Akademie věd Ukrajiny, ústav kybernetiky V.M.Gluškova
Place of publication
Kyjev
Event location
Kyjev
Event date
Jul 11, 2005
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—