Robust mixture of experts modeling using the t distribution.
Identifieur interne : 000B72 ( Main/Exploration ); précédent : 000B71; suivant : 000B73Robust mixture of experts modeling using the t distribution.
Auteurs : F. Chamroukhi [France]Source :
- Neural networks : the official journal of the International Neural Network Society [ 1879-2782 ] ; 2016.
Descripteurs français
- KwdFr :
- MESH :
English descriptors
- KwdEn :
- MESH :
Abstract
Mixture of Experts (MoE) is a popular framework for modeling heterogeneity in data for regression, classification, and clustering. For regression and cluster analyses of continuous data, MoE usually uses normal experts following the Gaussian distribution. However, for a set of data containing a group or groups of observations with heavy tails or atypical observations, the use of normal experts is unsuitable and can unduly affect the fit of the MoE model. We introduce a robust MoE modeling using the t distribution. The proposed t MoE (TMoE) deals with these issues regarding heavy-tailed and noisy data. We develop a dedicated expectation-maximization (EM) algorithm to estimate the parameters of the proposed model by monotonically maximizing the observed data log-likelihood. We describe how the presented model can be used in prediction and in model-based clustering of regression data. The proposed model is validated on numerical experiments carried out on simulated data, which show the effectiveness and the robustness of the proposed model in terms of modeling non-linear regression functions as well as in model-based clustering. Then, it is applied to the real-world data of tone perception for musical data analysis, and the one of temperature anomalies for the analysis of climate change data. The obtained results show the usefulness of the TMoE model for practical applications.
DOI: 10.1016/j.neunet.2016.03.002
PubMed: 27093693
Affiliations:
Links toward previous steps (curation, corpus...)
Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en">Robust mixture of experts modeling using the t distribution.</title><author><name sortKey="Chamroukhi, F" sort="Chamroukhi, F" uniqKey="Chamroukhi F" first="F" last="Chamroukhi">F. Chamroukhi</name><affiliation wicri:level="3"><nlm:affiliation>Université de Toulon, CNRS, LSIS, UMR 7296, 83957 La Garde, France; Aix Marseille Université, CNRS, ENSAM, LSIS, UMR 7296, 13397 Marseille, France; Laboratoire Paul Painlevé (LPP), UMR CNRS 8524, Université Lille 1, 59650 Villeneuve d'Ascq, France. Electronic address: faicel.chamroukhi@univ-tln.fr.</nlm:affiliation><country xml:lang="fr">France</country><wicri:regionArea>Université de Toulon, CNRS, LSIS, UMR 7296, 83957 La Garde, France; Aix Marseille Université, CNRS, ENSAM, LSIS, UMR 7296, 13397 Marseille, France; Laboratoire Paul Painlevé (LPP), UMR CNRS 8524, Université Lille 1, 59650 Villeneuve d'Ascq</wicri:regionArea><placeName><region type="region" nuts="2">Hauts-de-France</region><region type="old region" nuts="2">Nord-Pas-de-Calais</region><settlement type="city">Villeneuve d'Ascq</settlement></placeName></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">PubMed</idno><date when="2016">2016</date><idno type="RBID">pubmed:27093693</idno><idno type="pmid">27093693</idno><idno type="doi">10.1016/j.neunet.2016.03.002</idno><idno type="wicri:Area/Main/Corpus">000B86</idno><idno type="wicri:explorRef" wicri:stream="Main" wicri:step="Corpus" wicri:corpus="PubMed">000B86</idno><idno type="wicri:Area/Main/Curation">000B86</idno><idno type="wicri:explorRef" wicri:stream="Main" wicri:step="Curation">000B86</idno><idno type="wicri:Area/Main/Exploration">000B86</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en">Robust mixture of experts modeling using the t distribution.</title><author><name sortKey="Chamroukhi, F" sort="Chamroukhi, F" uniqKey="Chamroukhi F" first="F" last="Chamroukhi">F. Chamroukhi</name><affiliation wicri:level="3"><nlm:affiliation>Université de Toulon, CNRS, LSIS, UMR 7296, 83957 La Garde, France; Aix Marseille Université, CNRS, ENSAM, LSIS, UMR 7296, 13397 Marseille, France; Laboratoire Paul Painlevé (LPP), UMR CNRS 8524, Université Lille 1, 59650 Villeneuve d'Ascq, France. Electronic address: faicel.chamroukhi@univ-tln.fr.</nlm:affiliation><country xml:lang="fr">France</country><wicri:regionArea>Université de Toulon, CNRS, LSIS, UMR 7296, 83957 La Garde, France; Aix Marseille Université, CNRS, ENSAM, LSIS, UMR 7296, 13397 Marseille, France; Laboratoire Paul Painlevé (LPP), UMR CNRS 8524, Université Lille 1, 59650 Villeneuve d'Ascq</wicri:regionArea><placeName><region type="region" nuts="2">Hauts-de-France</region><region type="old region" nuts="2">Nord-Pas-de-Calais</region><settlement type="city">Villeneuve d'Ascq</settlement></placeName></affiliation></author></analytic><series><title level="j">Neural networks : the official journal of the International Neural Network Society</title><idno type="eISSN">1879-2782</idno><imprint><date when="2016" type="published">2016</date></imprint></series></biblStruct></sourceDesc></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Algorithms (MeSH)</term><term>Climate Change (MeSH)</term><term>Cluster Analysis (MeSH)</term><term>Models, Theoretical (MeSH)</term><term>Nonlinear Dynamics (MeSH)</term><term>Normal Distribution (MeSH)</term></keywords><keywords scheme="KwdFr" xml:lang="fr"><term>Algorithmes (MeSH)</term><term>Analyse de regroupements (MeSH)</term><term>Changement climatique (MeSH)</term><term>Dynamique non linéaire (MeSH)</term><term>Loi normale (MeSH)</term><term>Modèles théoriques (MeSH)</term></keywords><keywords scheme="MESH" xml:lang="en"><term>Algorithms</term><term>Climate Change</term><term>Cluster Analysis</term><term>Models, Theoretical</term><term>Nonlinear Dynamics</term><term>Normal Distribution</term></keywords><keywords scheme="MESH" xml:lang="fr"><term>Algorithmes</term><term>Analyse de regroupements</term><term>Changement climatique</term><term>Dynamique non linéaire</term><term>Loi normale</term><term>Modèles théoriques</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">Mixture of Experts (MoE) is a popular framework for modeling heterogeneity in data for regression, classification, and clustering. For regression and cluster analyses of continuous data, MoE usually uses normal experts following the Gaussian distribution. However, for a set of data containing a group or groups of observations with heavy tails or atypical observations, the use of normal experts is unsuitable and can unduly affect the fit of the MoE model. We introduce a robust MoE modeling using the t distribution. The proposed t MoE (TMoE) deals with these issues regarding heavy-tailed and noisy data. We develop a dedicated expectation-maximization (EM) algorithm to estimate the parameters of the proposed model by monotonically maximizing the observed data log-likelihood. We describe how the presented model can be used in prediction and in model-based clustering of regression data. The proposed model is validated on numerical experiments carried out on simulated data, which show the effectiveness and the robustness of the proposed model in terms of modeling non-linear regression functions as well as in model-based clustering. Then, it is applied to the real-world data of tone perception for musical data analysis, and the one of temperature anomalies for the analysis of climate change data. The obtained results show the usefulness of the TMoE model for practical applications. </div></front></TEI><pubmed><MedlineCitation Status="MEDLINE" IndexingMethod="Curated" Owner="NLM"><PMID Version="1">27093693</PMID><DateCompleted><Year>2017</Year><Month>04</Month><Day>19</Day></DateCompleted><DateRevised><Year>2018</Year><Month>12</Month><Day>02</Day></DateRevised><Article PubModel="Print-Electronic"><Journal><ISSN IssnType="Electronic">1879-2782</ISSN><JournalIssue CitedMedium="Internet"><Volume>79</Volume><PubDate><Year>2016</Year><Month>Jul</Month></PubDate></JournalIssue><Title>Neural networks : the official journal of the International Neural Network Society</Title><ISOAbbreviation>Neural Netw</ISOAbbreviation></Journal><ArticleTitle>Robust mixture of experts modeling using the t distribution.</ArticleTitle><Pagination><MedlinePgn>20-36</MedlinePgn></Pagination><ELocationID EIdType="doi" ValidYN="Y">10.1016/j.neunet.2016.03.002</ELocationID><ELocationID EIdType="pii" ValidYN="Y">S0893-6080(16)00043-5</ELocationID><Abstract><AbstractText>Mixture of Experts (MoE) is a popular framework for modeling heterogeneity in data for regression, classification, and clustering. For regression and cluster analyses of continuous data, MoE usually uses normal experts following the Gaussian distribution. However, for a set of data containing a group or groups of observations with heavy tails or atypical observations, the use of normal experts is unsuitable and can unduly affect the fit of the MoE model. We introduce a robust MoE modeling using the t distribution. The proposed t MoE (TMoE) deals with these issues regarding heavy-tailed and noisy data. We develop a dedicated expectation-maximization (EM) algorithm to estimate the parameters of the proposed model by monotonically maximizing the observed data log-likelihood. We describe how the presented model can be used in prediction and in model-based clustering of regression data. The proposed model is validated on numerical experiments carried out on simulated data, which show the effectiveness and the robustness of the proposed model in terms of modeling non-linear regression functions as well as in model-based clustering. Then, it is applied to the real-world data of tone perception for musical data analysis, and the one of temperature anomalies for the analysis of climate change data. The obtained results show the usefulness of the TMoE model for practical applications. </AbstractText><CopyrightInformation>Copyright © 2016 Elsevier Ltd. All rights reserved.</CopyrightInformation></Abstract><AuthorList CompleteYN="Y"><Author ValidYN="Y"><LastName>Chamroukhi</LastName><ForeName>F</ForeName><Initials>F</Initials><AffiliationInfo><Affiliation>Université de Toulon, CNRS, LSIS, UMR 7296, 83957 La Garde, France; Aix Marseille Université, CNRS, ENSAM, LSIS, UMR 7296, 13397 Marseille, France; Laboratoire Paul Painlevé (LPP), UMR CNRS 8524, Université Lille 1, 59650 Villeneuve d'Ascq, France. Electronic address: faicel.chamroukhi@univ-tln.fr.</Affiliation></AffiliationInfo></Author></AuthorList><Language>eng</Language><PublicationTypeList><PublicationType UI="D016428">Journal Article</PublicationType></PublicationTypeList><ArticleDate DateType="Electronic"><Year>2016</Year><Month>03</Month><Day>31</Day></ArticleDate></Article><MedlineJournalInfo><Country>United States</Country><MedlineTA>Neural Netw</MedlineTA><NlmUniqueID>8805018</NlmUniqueID><ISSNLinking>0893-6080</ISSNLinking></MedlineJournalInfo><CitationSubset>IM</CitationSubset><MeshHeadingList><MeshHeading><DescriptorName UI="D000465" MajorTopicYN="N">Algorithms</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D057231" MajorTopicYN="N">Climate Change</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D016000" MajorTopicYN="N">Cluster Analysis</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D008962" MajorTopicYN="Y">Models, Theoretical</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D017711" MajorTopicYN="Y">Nonlinear Dynamics</DescriptorName></MeshHeading><MeshHeading><DescriptorName UI="D016011" MajorTopicYN="Y">Normal Distribution</DescriptorName></MeshHeading></MeshHeadingList><KeywordList Owner="NOTNLM"><Keyword MajorTopicYN="N">EM algorithm</Keyword><Keyword MajorTopicYN="N">Mixture of experts</Keyword><Keyword MajorTopicYN="N">Model-based clustering</Keyword><Keyword MajorTopicYN="N">Non-linear regression</Keyword><Keyword MajorTopicYN="N">Robust modeling</Keyword><Keyword MajorTopicYN="N">distribution</Keyword></KeywordList></MedlineCitation><PubmedData><History><PubMedPubDate PubStatus="received"><Year>2015</Year><Month>09</Month><Day>17</Day></PubMedPubDate><PubMedPubDate PubStatus="revised"><Year>2016</Year><Month>01</Month><Day>23</Day></PubMedPubDate><PubMedPubDate PubStatus="accepted"><Year>2016</Year><Month>03</Month><Day>11</Day></PubMedPubDate><PubMedPubDate PubStatus="entrez"><Year>2016</Year><Month>4</Month><Day>20</Day><Hour>6</Hour><Minute>0</Minute></PubMedPubDate><PubMedPubDate PubStatus="pubmed"><Year>2016</Year><Month>4</Month><Day>20</Day><Hour>6</Hour><Minute>0</Minute></PubMedPubDate><PubMedPubDate PubStatus="medline"><Year>2017</Year><Month>4</Month><Day>20</Day><Hour>6</Hour><Minute>0</Minute></PubMedPubDate></History><PublicationStatus>ppublish</PublicationStatus><ArticleIdList><ArticleId IdType="pubmed">27093693</ArticleId><ArticleId IdType="pii">S0893-6080(16)00043-5</ArticleId><ArticleId IdType="doi">10.1016/j.neunet.2016.03.002</ArticleId></ArticleIdList></PubmedData></pubmed><affiliations><list><country><li>France</li></country><region><li>Hauts-de-France</li><li>Nord-Pas-de-Calais</li></region><settlement><li>Villeneuve d'Ascq</li></settlement></list><tree><country name="France"><region name="Hauts-de-France"><name sortKey="Chamroukhi, F" sort="Chamroukhi, F" uniqKey="Chamroukhi F" first="F" last="Chamroukhi">F. Chamroukhi</name></region></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Sante/explor/SanteMusiqueV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000B72 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000B72 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Sante
|area= SanteMusiqueV1
|flux= Main
|étape= Exploration
|type= RBID
|clé= pubmed:27093693
|texte= Robust mixture of experts modeling using the t distribution.
}}
Pour générer des pages wiki
HfdIndexSelect -h $EXPLOR_AREA/Data/Main/Exploration/RBID.i -Sk "pubmed:27093693" \
| HfdSelect -Kh $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd \
| NlmPubMed2Wicri -a SanteMusiqueV1
| This area was generated with Dilib version V0.6.38. |  |