Mining Frequent Synchronous Patterns based on Item Cover Similarity

Lade...
Vorschaubild
Dateien
Ezennaya-Gomez_2-1b2j5q1s61f8w3.pdf
Ezennaya-Gomez_2-1b2j5q1s61f8w3.pdfGröße: 693.45 KBDownloads: 258
Datum
2018
Autor:innen
Ezennaya-Gomez, Salatiel
Herausgeber:innen
Kontakt
ISSN der Zeitschrift
Electronic ISSN
ISBN
Bibliografische Daten
Verlag
Schriftenreihe
Auflagebezeichnung
DOI (zitierfähiger Link)
ArXiv-ID
Internationale Patentnummer
Link zur Lizenz
Angaben zur Forschungsförderung
Projekt
Open Access-Veröffentlichung
Open Access Gold
Core Facility der Universität Konstanz
Gesperrt bis
Titel in einer weiteren Sprache
Forschungsvorhaben
Organisationseinheiten
Zeitschriftenheft
Publikationstyp
Zeitschriftenartikel
Publikationsstatus
Published
Erschienen in
International Journal of Computational Intelligence Systems. 2018, 11(1), pp. 525-539. ISSN 1875-6891. eISSN 1875-6883. Available under: doi: 10.2991/ijcis.11.1.39
Zusammenfassung

In previous work we presented CoCoNAD (Continuous-time Closed Neuron Assembly Detection), a method to find significant synchronous patterns in parallel point processes with the goal to analyze parallel neural spike trains in neurobiology. A drawback of CoCoNAD and its accompanying methodology of pattern spectrum filtering (PSF) and pattern set reduction (PSR) is that it judges the (statistical) significance of a pattern only by the number of synchronous occurrences (support). However, the same number of occurrences can be significant for patterns consisting of items with a generally low occurrence rate, but explainable as a chance event for patterns consisting of items with a generally high occurrence rate, simply because more item occurrences produce more chance coincidences of items. In order to amend this drawback, we present in this paper an extension of the recently introduced CoCoNAD variant that is based on influence map overlap support (which takes both the number of synchronous events and the precision of synchrony into account), namely by transferring the idea of Jaccard item set mining to this setting: by basing pattern spectrum filtering upon item cover similarity measures, the number of coincidences is related to the item occurrence frequencies, which leads to an improved sensitivity for detecting synchronous events (or parallel episodes) in sequence data. We demonstrate the improved performance of our method by extensive experiments on artificial data sets.

Zusammenfassung in einer weiteren Sprache
Fachgebiet (DDC)
004 Informatik
Schlagwörter
graded synchrony, cover similarity, synchronous events, parallel episode, frequent pattern, pattern mining
Konferenz
Rezension
undefined / . - undefined, undefined
Zitieren
ISO 690EZENNAYA-GOMEZ, Salatiel, Christian BORGELT, 2018. Mining Frequent Synchronous Patterns based on Item Cover Similarity. In: International Journal of Computational Intelligence Systems. 2018, 11(1), pp. 525-539. ISSN 1875-6891. eISSN 1875-6883. Available under: doi: 10.2991/ijcis.11.1.39
BibTex
@article{EzennayaGomez2018Minin-45355,
  year={2018},
  doi={10.2991/ijcis.11.1.39},
  title={Mining Frequent Synchronous Patterns based on Item Cover Similarity},
  number={1},
  volume={11},
  issn={1875-6891},
  journal={International Journal of Computational Intelligence Systems},
  pages={525--539},
  author={Ezennaya-Gomez, Salatiel and Borgelt, Christian}
}
RDF
<rdf:RDF
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:bibo="http://purl.org/ontology/bibo/"
    xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#"
    xmlns:foaf="http://xmlns.com/foaf/0.1/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > 
  <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/45355">
    <dc:creator>Borgelt, Christian</dc:creator>
    <dc:contributor>Ezennaya-Gomez, Salatiel</dc:contributor>
    <foaf:homepage rdf:resource="http://localhost:8080/"/>
    <dcterms:title>Mining Frequent Synchronous Patterns based on Item Cover Similarity</dcterms:title>
    <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2019-03-07T13:23:57Z</dc:date>
    <dc:language>eng</dc:language>
    <dc:creator>Ezennaya-Gomez, Salatiel</dc:creator>
    <dspace:hasBitstream rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/45355/1/Ezennaya-Gomez_2-1b2j5q1s61f8w3.pdf"/>
    <dcterms:issued>2018</dcterms:issued>
    <dcterms:abstract xml:lang="eng">In previous work we presented CoCoNAD (Continuous-time Closed Neuron Assembly Detection), a method to find significant synchronous patterns in parallel point processes with the goal to analyze parallel neural spike trains in neurobiology. A drawback of CoCoNAD and its accompanying methodology of pattern spectrum filtering (PSF) and pattern set reduction (PSR) is that it judges the (statistical) significance of a pattern only by the number of synchronous occurrences (support). However, the same number of occurrences can be significant for patterns consisting of items with a generally low occurrence rate, but explainable as a chance event for patterns consisting of items with a generally high occurrence rate, simply because more item occurrences produce more chance coincidences of items. In order to amend this drawback, we present in this paper an extension of the recently introduced CoCoNAD variant that is based on influence map overlap support (which takes both the number of synchronous events and the precision of synchrony into account), namely by transferring the idea of Jaccard item set mining to this setting: by basing pattern spectrum filtering upon item cover similarity measures, the number of coincidences is related to the item occurrence frequencies, which leads to an improved sensitivity for detecting synchronous events (or parallel episodes) in sequence data. We demonstrate the improved performance of our method by extensive experiments on artificial data sets.</dcterms:abstract>
    <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2019-03-07T13:23:57Z</dcterms:available>
    <dc:contributor>Borgelt, Christian</dc:contributor>
    <dcterms:hasPart rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/45355/1/Ezennaya-Gomez_2-1b2j5q1s61f8w3.pdf"/>
    <dc:rights>Attribution-NonCommercial 4.0 International</dc:rights>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/36"/>
    <dcterms:rights rdf:resource="http://creativecommons.org/licenses/by-nc/4.0/"/>
    <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/45355"/>
  </rdf:Description>
</rdf:RDF>
Interner Vermerk
xmlui.Submission.submit.DescribeStep.inputForms.label.kops_note_fromSubmitter
Kontakt
URL der Originalveröffentl.
Prüfdatum der URL
Prüfungsdatum der Dissertation
Finanzierungsart
Kommentar zur Publikation
Allianzlizenz
Corresponding Authors der Uni Konstanz vorhanden
Internationale Co-Autor:innen
Universitätsbibliographie
Ja
Begutachtet
Unbekannt
Diese Publikation teilen