Knowledge extraction from neural networks for signal interpretation
Identifieur interne : 000C24 ( PascalFrancis/Checkpoint ); précédent : 000C23; suivant : 000C25Knowledge extraction from neural networks for signal interpretation
Auteurs : F. Alexandre [France] ; J.-F. Remm [France]Source :
Descripteurs français
- Pascal (Inist)
English descriptors
- KwdEn :
Abstract
Artificial neural networks have proved their ability to perform classification tasks. This ability is not satisfactory when expertise of the application domain is not available or when experts want to know more about hints that led to the decision. This leads presently to a great amount of work for knowledge or rule extraction from neural networks. In this paper, we propose a technique able to extract rules and to explain the functioning of the hidden layers of a multilayer perceptron. The first step consists in pruning the network with the classical OBD algorithm. Then, tightening of the sigmoidal transfer function can simply result in such knowledge extraction. This principle has been first tested on an application of signal interpretation in the radar domain.
Affiliations:
Links toward previous steps (curation, corpus...)
Links to Exploration step
Pascal:98-0230852Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Knowledge extraction from neural networks for signal interpretation</title><author><name sortKey="Alexandre, F" sort="Alexandre, F" uniqKey="Alexandre F" first="F." last="Alexandre">F. Alexandre</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN-INRIA, BP 239 </s1><s2>54506 Vandoeuvre </s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>2 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region" nuts="2">Grand Est</region><region type="old region" nuts="2">Lorraine (région)</region><settlement type="city">Vandoeuvre </settlement></placeName></affiliation></author><author><name sortKey="Remm, J F" sort="Remm, J F" uniqKey="Remm J" first="J.-F." last="Remm">J.-F. Remm</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN-INRIA, BP 239 </s1><s2>54506 Vandoeuvre </s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>2 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region" nuts="2">Grand Est</region><region type="old region" nuts="2">Lorraine (région)</region><settlement type="city">Vandoeuvre </settlement></placeName></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">98-0230852</idno><date when="1997">1997</date><idno type="stanalyst">PASCAL 98-0230852 INIST</idno><idno type="RBID">Pascal:98-0230852</idno><idno type="wicri:Area/PascalFrancis/Corpus">000C05</idno><idno type="wicri:Area/PascalFrancis/Curation">000C69</idno><idno type="wicri:Area/PascalFrancis/Checkpoint">000C24</idno><idno type="wicri:explorRef" wicri:stream="PascalFrancis" wicri:step="Checkpoint">000C24</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Knowledge extraction from neural networks for signal interpretation</title><author><name sortKey="Alexandre, F" sort="Alexandre, F" uniqKey="Alexandre F" first="F." last="Alexandre">F. Alexandre</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN-INRIA, BP 239 </s1><s2>54506 Vandoeuvre </s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>2 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region" nuts="2">Grand Est</region><region type="old region" nuts="2">Lorraine (région)</region><settlement type="city">Vandoeuvre </settlement></placeName></affiliation></author><author><name sortKey="Remm, J F" sort="Remm, J F" uniqKey="Remm J" first="J.-F." last="Remm">J.-F. Remm</name><affiliation wicri:level="3"><inist:fA14 i1="01"><s1>CRIN-INRIA, BP 239 </s1><s2>54506 Vandoeuvre </s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>2 aut.</sZ></inist:fA14><country>France</country><placeName><region type="region" nuts="2">Grand Est</region><region type="old region" nuts="2">Lorraine (région)</region><settlement type="city">Vandoeuvre </settlement></placeName></affiliation></author></analytic></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Information interpretation</term><term>Knowledge extraction</term><term>Multilayer perceptron</term><term>Neural network</term><term>Radar</term><term>Signal theory</term><term>Target</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>Radar</term><term>Réseau neuronal</term><term>Théorie signal</term><term>Interprétation information</term><term>Cible</term><term>Perceptron multicouche</term><term>Extraction connaissance</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Artificial neural networks have proved their ability to perform classification tasks. This ability is not satisfactory when expertise of the application domain is not available or when experts want to know more about hints that led to the decision. This leads presently to a great amount of work for knowledge or rule extraction from neural networks. In this paper, we propose a technique able to extract rules and to explain the functioning of the hidden layers of a multilayer perceptron. The first step consists in pruning the network with the classical OBD algorithm. Then, tightening of the sigmoidal transfer function can simply result in such knowledge extraction. This principle has been first tested on an application of signal interpretation in the radar domain.</div></front></TEI><inist><standard h6="B"><pA><fA08 i1="01" i2="1" l="ENG"><s1>Knowledge extraction from neural networks for signal interpretation</s1></fA08><fA09 i1="01" i2="1" l="ENG"><s1>ESANN '97 : European symposium on artificial neural networks : Bruges, 16-18 April 1997</s1></fA09><fA11 i1="01" i2="1"><s1>ALEXANDRE (F.)</s1></fA11><fA11 i1="02" i2="1"><s1>REMM (J.-F.)</s1></fA11><fA12 i1="01" i2="1"><s1>VERLEYSEN (Michel)</s1><s9>ed.</s9></fA12><fA14 i1="01"><s1>CRIN-INRIA, BP 239 </s1><s2>54506 Vandoeuvre </s2><s3>FRA</s3><sZ>1 aut.</sZ><sZ>2 aut.</sZ></fA14><fA20><s1>115-120</s1></fA20><fA21><s1>1997</s1></fA21><fA23 i1="01"><s0>ENG</s0></fA23><fA25 i1="01"><s1>D Facto</s1><s2>Brussels</s2></fA25><fA26 i1="01"><s0>2-9600049-7-3</s0></fA26><fA30 i1="01" i2="1" l="ENG"><s1>European symposium on artificial neural networks</s1><s2>5</s2><s3>Bruges BEL</s3><s4>1997-04-16</s4></fA30><fA43 i1="01"><s1>INIST</s1><s2>Y 31662</s2><s5>354000068109400190</s5></fA43><fA44><s0>0000</s0><s1>© 1998 INIST-CNRS. All rights reserved.</s1></fA44><fA45><s0>7 ref.</s0></fA45><fA47 i1="01" i2="1"><s0>98-0230852</s0></fA47><fA60><s1>C</s1></fA60><fA61><s0>A</s0></fA61><fA66 i1="01"><s0>BEL</s0></fA66><fC01 i1="01" l="ENG"><s0>Artificial neural networks have proved their ability to perform classification tasks. This ability is not satisfactory when expertise of the application domain is not available or when experts want to know more about hints that led to the decision. This leads presently to a great amount of work for knowledge or rule extraction from neural networks. In this paper, we propose a technique able to extract rules and to explain the functioning of the hidden layers of a multilayer perceptron. The first step consists in pruning the network with the classical OBD algorithm. Then, tightening of the sigmoidal transfer function can simply result in such knowledge extraction. This principle has been first tested on an application of signal interpretation in the radar domain.</s0></fC01><fC02 i1="01" i2="X"><s0>001D04B05</s0></fC02><fC02 i1="02" i2="X"><s0>001D02C06</s0></fC02><fC03 i1="01" i2="X" l="FRE"><s0>Radar</s0><s5>01</s5></fC03><fC03 i1="01" i2="X" l="ENG"><s0>Radar</s0><s5>01</s5></fC03><fC03 i1="01" i2="X" l="SPA"><s0>Radar</s0><s5>01</s5></fC03><fC03 i1="02" i2="X" l="FRE"><s0>Réseau neuronal</s0><s5>02</s5></fC03><fC03 i1="02" i2="X" l="ENG"><s0>Neural network</s0><s5>02</s5></fC03><fC03 i1="02" i2="X" l="SPA"><s0>Red neuronal</s0><s5>02</s5></fC03><fC03 i1="03" i2="X" l="FRE"><s0>Théorie signal</s0><s5>03</s5></fC03><fC03 i1="03" i2="X" l="ENG"><s0>Signal theory</s0><s5>03</s5></fC03><fC03 i1="03" i2="X" l="SPA"><s0>Teoría señal</s0><s5>03</s5></fC03><fC03 i1="04" i2="X" l="FRE"><s0>Interprétation information</s0><s5>04</s5></fC03><fC03 i1="04" i2="X" l="ENG"><s0>Information interpretation</s0><s5>04</s5></fC03><fC03 i1="04" i2="X" l="SPA"><s0>Interpretación información</s0><s5>04</s5></fC03><fC03 i1="05" i2="X" l="FRE"><s0>Cible</s0><s5>05</s5></fC03><fC03 i1="05" i2="X" l="ENG"><s0>Target</s0><s5>05</s5></fC03><fC03 i1="05" i2="X" l="SPA"><s0>Blanco</s0><s5>05</s5></fC03><fC03 i1="06" i2="X" l="FRE"><s0>Perceptron multicouche</s0><s4>CD</s4><s5>96</s5></fC03><fC03 i1="06" i2="X" l="ENG"><s0>Multilayer perceptron</s0><s4>CD</s4><s5>96</s5></fC03><fC03 i1="07" i2="X" l="FRE"><s0>Extraction connaissance</s0><s4>CD</s4><s5>97</s5></fC03><fC03 i1="07" i2="X" l="ENG"><s0>Knowledge extraction</s0><s4>CD</s4><s5>97</s5></fC03><fN21><s1>153</s1></fN21></pA></standard></inist><affiliations><list><country><li>France</li></country><region><li>Grand Est</li><li>Lorraine (région)</li></region><settlement><li>Vandoeuvre </li></settlement></list><tree><country name="France"><region name="Grand Est"><name sortKey="Alexandre, F" sort="Alexandre, F" uniqKey="Alexandre F" first="F." last="Alexandre">F. Alexandre</name></region><name sortKey="Remm, J F" sort="Remm, J F" uniqKey="Remm J" first="J.-F." last="Remm">J.-F. Remm</name></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/PascalFrancis/Checkpoint
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000C24 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/PascalFrancis/Checkpoint/biblio.hfd -nk 000C24 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Lorraine
|area= InforLorV4
|flux= PascalFrancis
|étape= Checkpoint
|type= RBID
|clé= Pascal:98-0230852
|texte= Knowledge extraction from neural networks for signal interpretation
}}
| This area was generated with Dilib version V0.6.33. |  |