FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3097436010> ?p ?o ?g. }

• W3097436010 abstract "The idea behind the emph{unsupervised} learning of emph{disentangled} representations is that real-world data is generated by a few explanatory factors of variation which can be recovered by unsupervised learning algorithms. In this paper, we provide a sober look at recent progress in the field and challenge some common assumptions. We first theoretically show that the unsupervised learning of representations is fundamentally impossible without inductive biases on both the models and the data. Then, we train over $14000$ models covering most prominent methods and evaluation metrics in a reproducible large-scale experimental study on eight data sets. We observe that while the different methods successfully enforce properties encouraged by the corresponding losses, well-disentangled models seemingly cannot be identified without supervision. Furthermore, different evaluation metrics do not always agree on what should be considered disentangled and exhibit systematic differences in the estimation. Finally, increased disentanglement does not seem to necessarily lead to a decreased sample complexity of learning for downstream tasks. Our results suggest that future work on disentanglement learning should be explicit about the role of inductive biases and (implicit) supervision, investigate concrete benefits of enforcing disentanglement of the learned representations, and consider a reproducible experimental setup covering several data sets." @default.
• W3097436010 created "2020-11-09" @default.
• W3097436010 creator A5018488407 @default.
• W3097436010 creator A5035416263 @default.
• W3097436010 creator A5041958651 @default.
• W3097436010 creator A5044005697 @default.
• W3097436010 creator A5044677020 @default.
• W3097436010 creator A5073157306 @default.
• W3097436010 creator A5089890773 @default.
• W3097436010 date "2020-10-27" @default.
• W3097436010 modified "2023-09-23" @default.
• W3097436010 title "A Sober Look at the Unsupervised Learning of Disentangled Representations and their Evaluation" @default.
• W3097436010 cites W1524326598 @default.
• W3097436010 cites W1574202531 @default.
• W3097436010 cites W1691728462 @default.
• W3097436010 cites W1782879880 @default.
• W3097436010 cites W1786904711 @default.
• W3097436010 cites W1912570122 @default.
• W3097436010 cites W1915328033 @default.
• W3097436010 cites W1959608418 @default.
• W3097436010 cites W1972673289 @default.
• W3097436010 cites W1994618660 @default.
• W3097436010 cites W2012762214 @default.
• W3097436010 cites W2095439994 @default.
• W3097436010 cites W2099741732 @default.
• W3097436010 cites W2101234009 @default.
• W3097436010 cites W2105728138 @default.
• W3097436010 cites W2124101779 @default.
• W3097436010 cites W2131024102 @default.
• W3097436010 cites W2134352661 @default.
• W3097436010 cites W2134557905 @default.
• W3097436010 cites W2144020560 @default.
• W3097436010 cites W2155183960 @default.
• W3097436010 cites W2157617585 @default.
• W3097436010 cites W2163922914 @default.
• W3097436010 cites W2166944917 @default.
• W3097436010 cites W2184218725 @default.
• W3097436010 cites W2188956040 @default.
• W3097436010 cites W2261396597 @default.
• W3097436010 cites W2281112906 @default.
• W3097436010 cites W2559823555 @default.
• W3097436010 cites W2613634265 @default.
• W3097436010 cites W2739693027 @default.
• W3097436010 cites W2753738274 @default.
• W3097436010 cites W2785961484 @default.
• W3097436010 cites W2789062658 @default.
• W3097436010 cites W2796704765 @default.
• W3097436010 cites W2801890059 @default.
• W3097436010 cites W2807255204 @default.
• W3097436010 cites W2808697642 @default.
• W3097436010 cites W2898909293 @default.
• W3097436010 cites W2902476877 @default.
• W3097436010 cites W2904849495 @default.
• W3097436010 cites W2920527457 @default.
• W3097436010 cites W2947918519 @default.
• W3097436010 cites W2962818303 @default.
• W3097436010 cites W2963045453 @default.
• W3097436010 cites W2963104724 @default.
• W3097436010 cites W2963117534 @default.
• W3097436010 cites W2963166838 @default.
• W3097436010 cites W2963201535 @default.
• W3097436010 cites W2963226019 @default.
• W3097436010 cites W2963264829 @default.
• W3097436010 cites W2963305465 @default.
• W3097436010 cites W2963366547 @default.
• W3097436010 cites W2963445340 @default.
• W3097436010 cites W2963618559 @default.
• W3097436010 cites W2963657860 @default.
• W3097436010 cites W2963762683 @default.
• W3097436010 cites W2963829960 @default.
• W3097436010 cites W2964127395 @default.
• W3097436010 cites W2964193753 @default.
• W3097436010 cites W2964342357 @default.
• W3097436010 cites W2964812973 @default.
• W3097436010 cites W2965354194 @default.
• W3097436010 cites W2966661 @default.
• W3097436010 cites W2969809866 @default.
• W3097436010 cites W2971127577 @default.
• W3097436010 cites W2989701728 @default.
• W3097436010 cites W2990376402 @default.
• W3097436010 cites W2996536863 @default.
• W3097436010 cites W2996554397 @default.
• W3097436010 cites W3035090613 @default.
• W3097436010 cites W3037805286 @default.
• W3097436010 cites W3042860357 @default.
• W3097436010 cites W2995089872 @default.
• W3097436010 hasPublicationYear "2020" @default.
• W3097436010 type Work @default.
• W3097436010 sameAs 3097436010 @default.
• W3097436010 citedByCount "2" @default.
• W3097436010 countsByYear W30974360102021 @default.
• W3097436010 crossrefType "posted-content" @default.
• W3097436010 hasAuthorship W3097436010A5018488407 @default.
• W3097436010 hasAuthorship W3097436010A5035416263 @default.
• W3097436010 hasAuthorship W3097436010A5041958651 @default.
• W3097436010 hasAuthorship W3097436010A5044005697 @default.
• W3097436010 hasAuthorship W3097436010A5044677020 @default.
• W3097436010 hasAuthorship W3097436010A5073157306 @default.
• W3097436010 hasAuthorship W3097436010A5089890773 @default.
• W3097436010 hasConcept C119857082 @default.

• next