Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.14279/10037
DC FieldValueLanguage
dc.contributor.authorPartaourides, Charalampos-
dc.contributor.authorChatzis, Sotirios P.-
dc.date.accessioned2017-04-25T09:44:13Z-
dc.date.available2017-04-25T09:44:13Z-
dc.date.issued2017-06-07-
dc.identifier.citationNeurocomputing, 2017, vol. 241, pp. 90-96en_US
dc.identifier.issn09252312-
dc.identifier.urihttps://hdl.handle.net/20.500.14279/10037-
dc.description.abstractAmortized variational inference, whereby the inferred latent variable posterior distributions are parameterized by means of neural network functions, has invigorated a new wave of innovation in the field of generative latent variable modeling, giving rise to the family of deep generative models (DGMs). Existing DGM formulations are based on the assumption of a symmetric Gaussian posterior over the model latent variables. This assumption, although mathematically convenient, can be well-expected to undermine the eventually obtained representation power, as it imposes apparent expressiveness limitations. Indeed, it has been recently shown that even some moderate increase in the latent variable posterior expressiveness, obtained by introducing an additional level of dependencies upon auxiliary (Gaussian) latent variables, can result in significant performance improvements in the context of semi-supervised learning tasks. Inspired from these advances, in this paper we examine whether a more potent increase in the expressiveness and representation power of modern DGMs can be achieved by completely relaxing their typical symmetric (Gaussian) latent variable posterior assumptions: Specifically, we consider DGMs with asymmetric posteriors, formulated as restricted multivariate skew-Normal (rMSN) distributions. We derive an efficient amortized variational inference algorithm for the proposed model, and exhibit its superiority over the current state-of-the-art in several semi-supervised learning benchmarks.en_US
dc.formatpdfen_US
dc.language.isoenen_US
dc.relation.ispartofNeurocomputingen_US
dc.rights© Elsevieren_US
dc.subjectDeep generative modelen_US
dc.subjectRestricted multivariate skew-Normal distributionen_US
dc.subjectSemi-supervised learningen_US
dc.subjectVariational inferenceen_US
dc.titleAsymmetric deep generative modelsen_US
dc.typeArticleen_US
dc.collaborationCyprus University of Technologyen_US
dc.subject.categoryElectrical Engineering - Electronic Engineering - Information Engineeringen_US
dc.journalsSubscriptionen_US
dc.countryCyprusen_US
dc.subject.fieldEngineering and Technologyen_US
dc.publicationPeer Revieweden_US
dc.identifier.doi10.1016/j.neucom.2017.02.028en_US
dc.relation.volume241en_US
cut.common.academicyear2016-2017en_US
dc.identifier.spage90en_US
dc.identifier.epage96en_US
item.fulltextNo Fulltext-
item.languageiso639-1en-
item.grantfulltextnone-
item.openairecristypehttp://purl.org/coar/resource_type/c_6501-
item.cerifentitytypePublications-
item.openairetypearticle-
crisitem.journal.journalissn0925-2312-
crisitem.journal.publisherElsevier-
crisitem.author.deptDepartment of Electrical Engineering, Computer Engineering and Informatics-
crisitem.author.deptDepartment of Electrical Engineering, Computer Engineering and Informatics-
crisitem.author.facultyFaculty of Engineering and Technology-
crisitem.author.facultyFaculty of Engineering and Technology-
crisitem.author.orcid0000-0002-8555-260X-
crisitem.author.orcid0000-0002-4956-4013-
crisitem.author.parentorgFaculty of Engineering and Technology-
crisitem.author.parentorgFaculty of Engineering and Technology-
Appears in Collections:Άρθρα/Articles
CORE Recommender
Show simple item record

SCOPUSTM   
Citations

13
checked on Nov 9, 2023

WEB OF SCIENCETM
Citations

9
Last Week
0
Last month
0
checked on Oct 29, 2023

Page view(s)

527
Last Week
0
Last month
33
checked on Mar 14, 2025

Google ScholarTM

Check

Altmetric


Items in KTISIS are protected by copyright, with all rights reserved, unless otherwise indicated.