Please use this identifier to cite or link to this item: http://dx.doi.org/10.14279/depositonce-10949
For citation please use:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorVidovic, Marina Marie-Claire-
dc.contributor.authorKloft, Marius-
dc.contributor.authorMüller, Klaus-Robert-
dc.contributor.authorGörnitz, Nico-
dc.date.accessioned2020-11-26T17:01:20Z-
dc.date.available2020-11-26T17:01:20Z-
dc.date.issued2017-03-27-
dc.identifier.urihttps://depositonce.tu-berlin.de/handle/11303/12075-
dc.identifier.urihttp://dx.doi.org/10.14279/depositonce-10949-
dc.description.abstractHigh prediction accuracies are not the only objective to consider when solving problems using machine learning. Instead, particular scientific applications require some explanation of the learned prediction function. For computational biology, positional oligomer importance matrices (POIMs) have been successfully applied to explain the decision of support vector machines (SVMs) using weighted-degree (WD) kernels. To extract relevant biological motifs from POIMs, the motifPOIM method has been devised and showed promising results on real-world data. Our contribution in this paper is twofold: as an extension to POIMs, we propose gPOIM, a general measure of feature importance for arbitrary learning machines and feature sets (including, but not limited to, SVMs and CNNs) and devise a sampling strategy for efficient computation. As a second contribution, we derive a convex formulation of motifPOIMs that leads to more reliable motif extraction from gPOIMs. Empirical evaluations confirm the usefulness of our approach on artificially generated data as well as on real-world datasets.en
dc.description.sponsorshipBMBF, 01IB15001B, ALICE II - Autonomes Lernen in komplexen Umgebungen 2 (Autonomous Learning in Complex Environments 2)en
dc.description.sponsorshipBMBF, 031L0023A, PREDICT - Umfassende Datenintegration zur Verbesserung onkologischer Therapien - Teilprojekt Aen
dc.description.sponsorshipBMBF, 031B0187B, Zuchtwert Mustererkennung in Hybridkulturarten (BreedPatH)-Teilprojekt Ben
dc.description.sponsorshipBMBF, 01IS14013A, Verbundprojekt: BBDC - Berliner Kompetenzzentrum für Big Dataen
dc.language.isoenen
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en
dc.subject.ddc006 Spezielle Computerverfahrenen
dc.subject.otherneural networken
dc.subject.othermachine learningen
dc.subject.otherpredictionen
dc.subject.othercomputational biologyen
dc.subject.othercomputationen
dc.subject.othernon-linear learning machinesen
dc.subject.otherML2Motifen
dc.titleML2Motif—Reliable extraction of discriminative sequence motifs from learning machinesen
dc.typeArticleen
tub.accessrights.dnbfreeen
tub.publisher.universityorinstitutionTechnische Universität Berlinen
dc.identifier.eissn1932-6203-
dc.type.versionpublishedVersionen
dcterms.bibliographicCitation.doi10.1371/journal.pone.0174392-
dcterms.bibliographicCitation.journaltitlePLOS ONEen
dcterms.bibliographicCitation.originalpublisherplaceSan Francisco, Calif.en
dcterms.bibliographicCitation.volume12-
dcterms.bibliographicCitation.originalpublishernamePublic Library of Science (PLOS)en
dcterms.bibliographicCitation.issue3-
dcterms.bibliographicCitation.articlenumbere0174392-
Appears in Collections:FG Maschinelles Lernen » Publications

Files in This Item:
journal.pone.0174392.pdf
Format: Adobe PDF | Size: 5.59 MB
DownloadShow Preview
Thumbnail

Item Export Bar

This item is licensed under a Creative Commons License Creative Commons