Show simple item record

dc.contributor.authorGane, Andreea
dc.contributor.authorJaakkola, Tommi S
dc.date.accessioned2021-01-19T15:13:42Z
dc.date.available2021-01-19T15:13:42Z
dc.date.issued2019-12
dc.date.submitted2019-10
dc.identifier.issn1049-5258
dc.identifier.urihttps://hdl.handle.net/1721.1/129438
dc.description.abstractReparameterization of variational auto-encoders with continuous random variables is an effective method for reducing the variance of their gradient estimates. In the discrete case, one can perform reparametrization using the Gumbel-Max trick, but the resulting objective relies on an arg max operation and is non-differentiable. In contrast to previous works which resort to softmax-based relaxations, we propose to optimize it directly by applying the direct loss minimization approach. Our proposal extends naturally to structured discrete latent variable models when evaluating the arg max operation is tractable. We demonstrate empirically the effectiveness of the direct loss minimization technique in variational autoencoders with both unstructured and structured discrete latent variables.en_US
dc.language.isoen
dc.publisherMorgan Kaufmann Publishersen_US
dc.rightsArticle is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.en_US
dc.sourceNeural Information Processing Systems (NIPS)en_US
dc.titleDirect optimization through arg max for discrete variational auto-encoderen_US
dc.typeArticleen_US
dc.identifier.citationLorberbom, Guy et al. “Direct optimization through arg max for discrete variational auto-encoder.” 33rd Conference on Neural Information Processing Systems, December 2019, Vancouver, Canada, Morgan Kaufmann Publishers, 2019. © 2019 The Author(s)en_US
dc.contributor.departmentMassachusetts Institute of Technology. Department of Electrical Engineering and Computer Scienceen_US
dc.relation.journal33rd Conference on Neural Information Processing Systems (NeurIPS 2019)en_US
dc.eprint.versionFinal published versionen_US
dc.type.urihttp://purl.org/eprint/type/ConferencePaperen_US
eprint.statushttp://purl.org/eprint/status/NonPeerRevieweden_US
dc.date.updated2020-12-21T15:49:12Z
dspace.orderedauthorsLorberbom, G; Gane, A; Jaakkola, T; Hazan, Ten_US
dspace.date.submission2020-12-21T15:49:15Z
mit.journal.volume32en_US
mit.licensePUBLISHER_POLICY


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record