View source for Publications:Selecting neural networks for a committee decision

<div style='display: none'>
== Do not edit this section ==
</div>
{{PublicationSetupTemplate|Author=Antanas Verikas, Arunas Lipnickas, Kerstin Malmqvist
|PID=285853
|Name=Verikas, Antanas [av] (Högskolan i Halmstad [2804], Sektionen för Informationsvetenskap, Data– och Elektroteknik (IDE) [3905], Halmstad Embedded and Intelligent Systems Research (EIS) [3938]);Lipnickas, Arunas (Department of Applied Electronics, Kaunas University of Technology, Lithuania);Malmqvist, Kerstin (Högskolan i Halmstad [2804], Sektionen för Informationsvetenskap, Data– och Elektroteknik (IDE) [3905], Halmstad Embedded and Intelligent Systems Research (EIS) [3938], Intelligenta system (IS-lab) [3941])
|Title=Selecting neural networks for a committee decision
|PublicationType=Journal Paper
|ContentType=Refereegranskat
|Language=eng
|Journal=International Journal of Neural Systems
|JournalISSN=0129-0657
|Status=published
|Volume=12
|Issue=5
|HostPublication=
|Conference=
|StartPage=351
|EndPage=361
|Year=2002
|Edition=
|Pages=
|City=Singapore
|Publisher=World Scientific
|Series=
|SeriesISSN=
|ISBN=
|Urls=
|ISRN=
|DOI=http://dx.doi.org/10.1142/S0129065702001229
|ISI=
|PMID=12424806
|ScopusId=2-s2.0-2342502764
|NBN=urn:nbn:se:hh:diva-3541
|LocalId=
|ArchiveNumber=
|Keywords=Artificial Intelligence;Computer Simulation;Neural Networks (Computer);Neural network committee;Decision fusion;Neural network selection
|Categories=Datavetenskap (datalogi) (10201)
|ResearchSubjects=
|Projects=
|Notes=
|Abstract=<p>To improve recognition results, decisions of multiple neural networks can be aggregated into a committee decision. In contrast to the ordinary approach of utilizing all neural networks available to make a committee decision, we propose creating adaptive committees, which are specific for each input data point. A prediction network is used to identify classification neural networks to be fused for making a committee decision about a given input data point. The jth output value of the prediction network expresses the expectation level that the jth classification neural network will make a correct decision about the class label of a given input data point. The proposed technique is tested in three aggregation schemes, namely majority vote, averaging, and aggregation by the median rule and compared with the ordinary neural networks fusion approach. The effectiveness of the approach is demonstrated on two artificial and three real data sets.</p>
|Opponents=
|Supervisors=
|Examiners=
|Patent=
|ThesisLevel=
|Credits=
|Programme=
|Subject=
|Uppsok=
|DefencePlace=
|DefenceLanguage=
|DefenceDate=
|CreatedDate=2009-12-01
|PublicationDate=2010-01-13
|LastUpdated=2013-06-11
|diva=http://hh.diva-portal.org/smash/record.jsf?searchId=1&pid=diva2:285853}}
<div style='display: none'>

== Keep all hand-made modifications below ==
</div>{{PublicationDisplayTemplate}}