Ensemble supervised learning for genomic selectionJubair • 被引用 4 次Pattern recognition in multivariate time series: towards …Kapp • 被引用 7 次Artificial neural networksNeural networkssimple and effective stochastic neural networkslearning stochastic feedforward neural networksbayesian neural networkStochastic Neural Networks - Microsoft Research
文章推薦指數： 80 %
In the stochastic neural network project we aim to build the next generation of deep learning models which are more data-efficient and can enable machines ... SkiptoHeader SkiptoSearch SkiptoContent SkiptoFooter Skiptomaincontent StochasticNeuralNetworks Overview Publications MicrosoftResearchblog Willmachinesonedaybeascreativeashumans? Whenwewritealetter,haveaconversation,ordrawapicture,weexerciseauniquelyhumanskillbycreatingcomplexartifactsthatembodyinformation.CurrentAItechnologycannotyetmatchhumanabilityinthisareabecauseitfailstohavethesameunderstandingoftheworldintermsofindependentcausalfactors.WithoutsuchunderstandingAIsystemscannotcomposeinformationintherichwaysahumancould. Instead,currentsuccessfulAItechnologyrequiresalargeamountofsuperviseddatainordertolearndifferentplausiblecombinationsofindependentvariations. Inthestochasticneuralnetworkprojectweaimtobuildthenextgenerationofdeeplearningmodelswhicharemoredata-efficientandcanenablemachinestolearnmoreefficientlyandeventuallytobetrulycreative. ResearchDirection Researchinthestochasticneuralnetworksprojectaddressesthisresearchchallengealongthreelines: Developingnovelalgorithmsfordeepprobabilisticmodels; Learningdisentangledrepresentationsofcomplexdata; Applicationsofdeepprobabilisticmodelstoapplications. NovelAlgorithmsforDeepProbabilisticModels Weanalyzeandimproveimportantalgorithmsforgenerativemodelssuchasmethodsforgenerativeadversarialnetworks,modelsbasedonintegralprobabilitymetrics,andvariationalautoencoders. Variationalcharacterizationoff-divergencesinf-GAN. LearningDisentangledRepresentationsofComplexData Wedevelopnovelmodelstolearnlatentfactorrepresentationsofcomplexdata,includingmodelsthatcanlearndisentangledrepresentationsusingminimalsupervision. Disentangledrepresentationoffaces,allowingstyletransferbetweenpeople. ApplicationsofDeepProbabilisticModelstoApplications Weapplydeepprobabilisticmodelstochallengingapplicationswithheteroscedasticuncertainty;applicationareasincludecomputervision,reinforcementlearning,andnaturallanguagemodels. ResNetGANwithandwithoutstabilization. People People SebastianNowozin PartnerResearchManager Learnmore RyotaTomioka PrincipalResearchManager Learnmore KatjaHofmann SeniorPrincipalResearcher Learnmore Followus: FollowonTwitter LikeonFacebook SubscribeonYoutube FollowonInstagram SubscribetoourRSSfeed Sharethispage: ShareonTwitter ShareonFacebook ShareonLinkedIn ShareonReddit
延伸文章資訊
- 1Stochastic Neural Networks with Infinite Width are Deterministic
Specifically, we prove that as the width of an optimized stochastic neural network tends to infin...
- 2[1910.09763] Stochastic Feedforward Neural Networks - arXiv
In contrast to deterministic networks, which represent mappings from a set of inputs to a set of ...
- 3Simple and Effective Stochastic Neural Networks - Association ...
Stochastic neural networks (SNNs) are currently topical, with several paradigms being actively in...
- 4Are Neural Networks Stochastic Or Deterministic? - Phoenixite
- 5Probability Matching via Deterministic Neural Networks - CEUR-WS
