The combination of multiple classifiers by a neural network approach
Identifieur interne : 000948 ( PascalFrancis/Curation ); précédent : 000947; suivant : 000949The combination of multiple classifiers by a neural network approach
Auteurs : Y. S. Huang [Canada] ; K. Liu ; C. Y. SuenSource :
- International journal of pattern recognition and artificial intelligence [ 0218-0014 ] ; 1995.
Descripteurs français
- Pascal (Inist)
- Wicri :
- topic : Base de données.
English descriptors
- KwdEn :
Abstract
Due to different writing styles and various kinds of noise, the recognition of handwritten numerals is an extremely complicated problem. Recently, a new trend has emerged to tackle this problem by the use of multiple classifiers. This method combines individual classification decisions to derive the final decisions. This is called "Combination of Multiple Classifiers" (CME). In this paper, a novel approach to CME is developed and discussed in detail. It contains two steps: data transformation and data classification. In data transformation, the output values of each classifier are first transformed into a form of likeness measurement. The larger a likeness measurement is, the more probable the corresponding class has the input. In data classification, neural networks have been found very suitable to aggregate the transformed output to produce the final classification decisions. Some strategies for further improving the performance of neural networks have also been proposed in this paper. Experiments with several data transformation functions and data classification approaches have been performed on a large number of handwritten samples. The best result among them is achieved by using both the proposed data transformation function and the multi-layer perceptron neural net, which increased the recognition rate of three individual classifications considerably.
pA |
|
---|
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: Pour aller vers cette notice dans l'étape Curation :000A51
Links to Exploration step
Pascal:95-0511946Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">The combination of multiple classifiers by a neural network approach</title>
<author><name sortKey="Huang, Y S" sort="Huang, Y S" uniqKey="Huang Y" first="Y. S." last="Huang">Y. S. Huang</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Concordia univ., cent. pattern recognition machine intelligence</s1>
<s2>Montréal PQ H3G 1M8</s2>
<s3>CAN</s3>
</inist:fA14>
<country>Canada</country>
</affiliation>
</author>
<author><name sortKey="Liu, K" sort="Liu, K" uniqKey="Liu K" first="K." last="Liu">K. Liu</name>
</author>
<author><name sortKey="Suen, C Y" sort="Suen, C Y" uniqKey="Suen C" first="C. Y." last="Suen">C. Y. Suen</name>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">INIST</idno>
<idno type="inist">95-0511946</idno>
<date when="1995">1995</date>
<idno type="stanalyst">PASCAL 95-0511946 INIST</idno>
<idno type="RBID">Pascal:95-0511946</idno>
<idno type="wicri:Area/PascalFrancis/Corpus">000A51</idno>
<idno type="wicri:Area/PascalFrancis/Curation">000948</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">The combination of multiple classifiers by a neural network approach</title>
<author><name sortKey="Huang, Y S" sort="Huang, Y S" uniqKey="Huang Y" first="Y. S." last="Huang">Y. S. Huang</name>
<affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Concordia univ., cent. pattern recognition machine intelligence</s1>
<s2>Montréal PQ H3G 1M8</s2>
<s3>CAN</s3>
</inist:fA14>
<country>Canada</country>
</affiliation>
</author>
<author><name sortKey="Liu, K" sort="Liu, K" uniqKey="Liu K" first="K." last="Liu">K. Liu</name>
</author>
<author><name sortKey="Suen, C Y" sort="Suen, C Y" uniqKey="Suen C" first="C. Y." last="Suen">C. Y. Suen</name>
</author>
</analytic>
<series><title level="j" type="main">International journal of pattern recognition and artificial intelligence</title>
<title level="j" type="abbreviated">Int. j. pattern recogn. artif. intell.</title>
<idno type="ISSN">0218-0014</idno>
<imprint><date when="1995">1995</date>
</imprint>
</series>
</biblStruct>
</sourceDesc>
<seriesStmt><title level="j" type="main">International journal of pattern recognition and artificial intelligence</title>
<title level="j" type="abbreviated">Int. j. pattern recogn. artif. intell.</title>
<idno type="ISSN">0218-0014</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Algorithm performance</term>
<term>Character recognition</term>
<term>Classifiers</term>
<term>Database</term>
<term>Decision rule</term>
<term>Gradient method</term>
<term>Iterative methods</term>
<term>Manuscript character</term>
</keywords>
<keywords scheme="Pascal" xml:lang="fr"><term>Reconnaissance caractère</term>
<term>Classificateur</term>
<term>Caractère manuscrit</term>
<term>Performance algorithme</term>
<term>Méthode itérative</term>
<term>Méthode gradient</term>
<term>Règle décision</term>
<term>Base donnée</term>
<term>Unconstrained handwriting recognition</term>
<term>OCR</term>
<term>Multi layer perceptron</term>
<term>Generalized delta rule</term>
</keywords>
<keywords scheme="Wicri" type="topic" xml:lang="fr"><term>Base de données</term>
</keywords>
</textClass>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">Due to different writing styles and various kinds of noise, the recognition of handwritten numerals is an extremely complicated problem. Recently, a new trend has emerged to tackle this problem by the use of multiple classifiers. This method combines individual classification decisions to derive the final decisions. This is called "Combination of Multiple Classifiers" (CME). In this paper, a novel approach to CME is developed and discussed in detail. It contains two steps: data transformation and data classification. In data transformation, the output values of each classifier are first transformed into a form of likeness measurement. The larger a likeness measurement is, the more probable the corresponding class has the input. In data classification, neural networks have been found very suitable to aggregate the transformed output to produce the final classification decisions. Some strategies for further improving the performance of neural networks have also been proposed in this paper. Experiments with several data transformation functions and data classification approaches have been performed on a large number of handwritten samples. The best result among them is achieved by using both the proposed data transformation function and the multi-layer perceptron neural net, which increased the recognition rate of three individual classifications considerably.</div>
</front>
</TEI>
<inist><standard h6="B"><pA><fA01 i1="01" i2="1"><s0>0218-0014</s0>
</fA01>
<fA03 i2="1"><s0>Int. j. pattern recogn. artif. intell.</s0>
</fA03>
<fA05><s2>9</s2>
</fA05>
<fA06><s2>3</s2>
</fA06>
<fA08 i1="01" i2="1" l="ENG"><s1>The combination of multiple classifiers by a neural network approach</s1>
</fA08>
<fA11 i1="01" i2="1"><s1>HUANG (Y. S.)</s1>
</fA11>
<fA11 i1="02" i2="1"><s1>LIU (K.)</s1>
</fA11>
<fA11 i1="03" i2="1"><s1>SUEN (C. Y.)</s1>
</fA11>
<fA14 i1="01"><s1>Concordia univ., cent. pattern recognition machine intelligence</s1>
<s2>Montréal PQ H3G 1M8</s2>
<s3>CAN</s3>
</fA14>
<fA20><s1>579-597</s1>
</fA20>
<fA21><s1>1995</s1>
</fA21>
<fA23 i1="01"><s0>ENG</s0>
</fA23>
<fA43 i1="01"><s1>INIST</s1>
<s2>22088</s2>
<s5>354000054571070070</s5>
</fA43>
<fA44><s0>0000</s0>
</fA44>
<fA45><s0>21 ref.</s0>
</fA45>
<fA47 i1="01" i2="1"><s0>95-0511946</s0>
</fA47>
<fA60><s1>P</s1>
</fA60>
<fA61><s0>A</s0>
</fA61>
<fA64 i1="01" i2="1"><s0>International journal of pattern recognition and artificial intelligence</s0>
</fA64>
<fA66 i1="01"><s0>SGP</s0>
</fA66>
<fC01 i1="01" l="ENG"><s0>Due to different writing styles and various kinds of noise, the recognition of handwritten numerals is an extremely complicated problem. Recently, a new trend has emerged to tackle this problem by the use of multiple classifiers. This method combines individual classification decisions to derive the final decisions. This is called "Combination of Multiple Classifiers" (CME). In this paper, a novel approach to CME is developed and discussed in detail. It contains two steps: data transformation and data classification. In data transformation, the output values of each classifier are first transformed into a form of likeness measurement. The larger a likeness measurement is, the more probable the corresponding class has the input. In data classification, neural networks have been found very suitable to aggregate the transformed output to produce the final classification decisions. Some strategies for further improving the performance of neural networks have also been proposed in this paper. Experiments with several data transformation functions and data classification approaches have been performed on a large number of handwritten samples. The best result among them is achieved by using both the proposed data transformation function and the multi-layer perceptron neural net, which increased the recognition rate of three individual classifications considerably.</s0>
</fC01>
<fC02 i1="01" i2="1"><s0>001D02C03</s0>
</fC02>
<fC03 i1="01" i2="1" l="FRE"><s0>Reconnaissance caractère</s0>
<s3>P</s3>
<s5>01</s5>
</fC03>
<fC03 i1="01" i2="1" l="ENG"><s0>Character recognition</s0>
<s3>P</s3>
<s5>01</s5>
</fC03>
<fC03 i1="02" i2="1" l="FRE"><s0>Classificateur</s0>
<s5>02</s5>
</fC03>
<fC03 i1="02" i2="1" l="ENG"><s0>Classifiers</s0>
<s5>02</s5>
</fC03>
<fC03 i1="03" i2="X" l="FRE"><s0>Caractère manuscrit</s0>
<s5>16</s5>
</fC03>
<fC03 i1="03" i2="X" l="ENG"><s0>Manuscript character</s0>
<s5>16</s5>
</fC03>
<fC03 i1="03" i2="X" l="SPA"><s0>Carácter manuscrito</s0>
<s5>16</s5>
</fC03>
<fC03 i1="04" i2="X" l="FRE"><s0>Performance algorithme</s0>
<s5>17</s5>
</fC03>
<fC03 i1="04" i2="X" l="ENG"><s0>Algorithm performance</s0>
<s5>17</s5>
</fC03>
<fC03 i1="04" i2="X" l="SPA"><s0>Resultado algoritmo</s0>
<s5>17</s5>
</fC03>
<fC03 i1="05" i2="1" l="FRE"><s0>Méthode itérative</s0>
<s5>18</s5>
</fC03>
<fC03 i1="05" i2="1" l="ENG"><s0>Iterative methods</s0>
<s5>18</s5>
</fC03>
<fC03 i1="06" i2="X" l="FRE"><s0>Méthode gradient</s0>
<s5>19</s5>
</fC03>
<fC03 i1="06" i2="X" l="ENG"><s0>Gradient method</s0>
<s5>19</s5>
</fC03>
<fC03 i1="06" i2="X" l="SPA"><s0>Método gradiente</s0>
<s5>19</s5>
</fC03>
<fC03 i1="07" i2="X" l="FRE"><s0>Règle décision</s0>
<s5>20</s5>
</fC03>
<fC03 i1="07" i2="X" l="ENG"><s0>Decision rule</s0>
<s5>20</s5>
</fC03>
<fC03 i1="07" i2="X" l="SPA"><s0>Regla decisión</s0>
<s5>20</s5>
</fC03>
<fC03 i1="08" i2="X" l="FRE"><s0>Base donnée</s0>
<s5>21</s5>
</fC03>
<fC03 i1="08" i2="X" l="ENG"><s0>Database</s0>
<s5>21</s5>
</fC03>
<fC03 i1="08" i2="X" l="SPA"><s0>Base dato</s0>
<s5>21</s5>
</fC03>
<fC03 i1="09" i2="1" l="FRE"><s0>Unconstrained handwriting recognition</s0>
<s4>INC</s4>
<s5>72</s5>
</fC03>
<fC03 i1="10" i2="1" l="FRE"><s0>OCR</s0>
<s4>INC</s4>
<s5>73</s5>
</fC03>
<fC03 i1="11" i2="1" l="FRE"><s0>Multi layer perceptron</s0>
<s4>INC</s4>
<s5>74</s5>
</fC03>
<fC03 i1="12" i2="1" l="FRE"><s0>Generalized delta rule</s0>
<s4>INC</s4>
<s5>75</s5>
</fC03>
<fN21><s1>289</s1>
</fN21>
</pA>
</standard>
</inist>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/OcrV1/Data/PascalFrancis/Curation
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000948 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/PascalFrancis/Curation/biblio.hfd -nk 000948 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Ticri/CIDE |area= OcrV1 |flux= PascalFrancis |étape= Curation |type= RBID |clé= Pascal:95-0511946 |texte= The combination of multiple classifiers by a neural network approach }}
![]() | This area was generated with Dilib version V0.6.32. | ![]() |