Show simple item record

dc.contributor.authorDaw, Arkaen
dc.contributor.authorMaruf, M.en
dc.contributor.authorKarpatne, Anujen
dc.date.accessioned2021-04-26T16:40:49Zen
dc.date.available2021-04-26T16:40:49Zen
dc.date.issued2020-12-11en
dc.identifier.urihttp://hdl.handle.net/10919/103125en
dc.description.abstractWe propose a novel method of incorporating physical knowledge as an additional input to the discriminator of a conditional Generative Adversarial Net (cGAN). Our proposed approach, termed as Physics-informed Discriminator for cGAN (cGAN-PID), is more aligned to the adversarial learning idea of cGAN as opposed to existing methods on incorporating physical knowledge in GANs by adding physics based loss functions as additional terms in the optimization objective of GAN. We evaluate the performance of our model on two toy datasets and demonstrate that our proposed cGAN-PID can be used as an alternative to the existing techniques.en
dc.format.mimetypeapplication/pdfen
dc.language.isoenen
dc.publisherNeurIPSen
dc.relation.ispartofMachine Learning and Physical Sciences workshop at Neural Information Processing Systems (NeurIPS) 2020.en
dc.rightsAttribution 4.0 Internationalen
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/en
dc.subjectMachine Learningen
dc.subjectGenerative Adversarial Networksen
dc.subjectPhysics Informed Neural Networksen
dc.subjectuncertainty quantificationen
dc.subjecttheory-guided data scienceen
dc.titlePhysics-Informed Discriminator (PID) for Conditional Generative Adversarial Netsen
dc.typeConference proceedingen
dc.contributor.departmentComputer Scienceen
dc.type.dcmitypeTexten


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution 4.0 International
License: Attribution 4.0 International