Repository logo
 
Publication

On the Use of Kullback–Leibler Divergence for Kernel Selection and Interpretation in Variational Autoencoders for Feature Creation

2023Journal article Open access
Simple item page
dc.contributor.authorMendonça, Fábio
dc.contributor.authorMostafa, Sheikh Shanawaz
dc.contributor.authorDias, Fernando Morgado
dc.contributor.authorRavelo-García, Antonio G.
dc.date.accessioned2024-02-19T11:53:02Z
dc.date.available2024-02-19T11:53:02Z
dc.date.issued2023
dc.description.abstractThis study presents a novel approach for kernel selection based on Kullback–Leibler divergence in variational autoencoders using features generated by the convolutional encoder. The proposed methodology focuses on identifying the most relevant subset of latent variables to reduce the model’s parameters. Each latent variable is sampled from the distribution associated with a single kernel of the last encoder’s convolutional layer, resulting in an individual distribution for each kernel. Relevant features are selected from the sampled latent variables to perform kernel selection, which filters out uninformative features and, consequently, unnecessary kernels. Both the proposed filter method and the sequential feature selection (standard wrapper method) were examined for feature selection. Particularly, the filter method evaluates the Kullback–Leibler divergence between all kernels’ distributions and hypothesizes that similar kernels can be discarded as they do not convey relevant information. This hypothesis was confirmed through the experiments performed on four standard datasets, where it was observed that the number of kernels can be reduced without meaningfully affecting the performance. This analysis was based on the accuracy of the model when the selected kernels fed a probabilistic classifier and the feature-based similarity index to appraise the quality of the reconstructed images when the variational autoencoder only uses the selected kernels. Therefore, the proposed methodology guides the reduction of the number of parameters of the model, making it suitable for developing applications for resource-constrained devices.pt_PT
dc.description.versioninfo:eu-repo/semantics/publishedVersionpt_PT
dc.identifier.citationMendonça, F.; Mostafa, S.S.; Morgado-Dias, F.; Ravelo-García, A.G. On the Use of Kullback–Leibler Divergence for Kernel Selection and Interpretation in Variational Autoencoders for Feature Creation. Information 2023, 14, 571. https:// doi.org/10.3390/info14100571pt_PT
dc.identifier.doi10.3390/info14100571pt_PT
dc.identifier.urihttp://hdl.handle.net/10400.13/5563
dc.language.isoengpt_PT
dc.peerreviewedyespt_PT
dc.publisherMDPIpt_PT
dc.relationLaboratory of Robotics and Engineering Systems
dc.relationLaboratory for Robotics and Engineering Systems
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/pt_PT
dc.subjectConvolutional neural networkpt_PT
dc.subjectFeature selectionpt_PT
dc.subjectLatent variablespt_PT
dc.subjectProbabilistic classifierpt_PT
dc.subjectVariational autoencoderpt_PT
dc.subject.pt_PT
dc.subjectFaculdade de Ciências Exatas e da Engenhariapt_PT
dc.titleOn the Use of Kullback–Leibler Divergence for Kernel Selection and Interpretation in Variational Autoencoders for Feature Creationpt_PT
dc.typejournal article
dspace.entity.typePublication
oaire.awardTitleLaboratory of Robotics and Engineering Systems
oaire.awardTitleLaboratory for Robotics and Engineering Systems
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UIDB%2F50009%2F2020/PT
oaire.awardURIinfo:eu-repo/grantAgreement/FCT/6817 - DCRRNI ID/UID%2FEEA%2F50009%2F2019/PT
oaire.citation.issue10pt_PT
oaire.citation.startPage571pt_PT
oaire.citation.titleInformationpt_PT
oaire.citation.volume14pt_PT
oaire.fundingStream6817 - DCRRNI ID
oaire.fundingStream6817 - DCRRNI ID
person.familyNameSilva Mendonça
person.familyNameMostafa
person.familyNameMorgado-Dias
person.familyNameRavelo-García
person.givenNameFábio Rúben
person.givenNameSheikh Shanawaz
person.givenNameFernando
person.givenNameAntonio G.
person.identifier34497
person.identifier.ciencia-id7F1E-8AE9-3098
person.identifier.ciencia-idEE14-BEB3-F82B
person.identifier.ciencia-id7B14-DF07-AA6D
person.identifier.orcid0000-0002-5107-3248
person.identifier.orcid0000-0002-7677-0971
person.identifier.orcid0000-0001-7334-3993
person.identifier.orcid0000-0002-8512-965X
person.identifier.ridN-9228-2015
person.identifier.scopus-author-id55489640900
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.identifierhttp://doi.org/10.13039/501100001871
project.funder.nameFundação para a Ciência e a Tecnologia
project.funder.nameFundação para a Ciência e a Tecnologia
rcaap.rightsopenAccesspt_PT
rcaap.typearticlept_PT
relation.isAuthorOfPublication0dbad4bf-3a0c-4c43-8022-42199a5e09c0
relation.isAuthorOfPublicationf90aafd0-eedb-47ea-945a-40b1c1fe802a
relation.isAuthorOfPublication042f7593-c6ca-4553-8f0e-12ccf17018be
relation.isAuthorOfPublication866dcd15-c600-4e6a-8a06-48c3d58a69a5
relation.isAuthorOfPublication.latestForDiscovery0dbad4bf-3a0c-4c43-8022-42199a5e09c0
relation.isProjectOfPublicationc0352e81-99e6-4923-99ed-074df09e4db0
relation.isProjectOfPublicationeda3c844-853e-4c97-83da-99235180d98c
relation.isProjectOfPublication.latestForDiscoveryeda3c844-853e-4c97-83da-99235180d98c

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
On the Use of Kullback–Leibler Divergence for Kernel Selection and Interpretation in Variational Autoencoders for Feature Creation.pdf
Size:
3.95 MB
Format:
Adobe Portable Document Format
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Artigos em revistas internacionais