Matches in SemOpenAlex for { <https://semopenalex.org/work/W4376982954> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W4376982954 endingPage "3772" @default.
- W4376982954 startingPage "3749" @default.
- W4376982954 abstract "It can be difficult to assess the quality of a fitted model when facing unsupervised learning problems. Latent variable models, such as variation autoencoders and Gaussian mixture models, are often trained with likelihood-based approaches. In scope of Goodhart's law, when a metric becomes a target it ceases to be a good metric and therefore we should not use likelihood to assess the quality of the fit of these models. The solution we propose is a new metric for model comparison or regularization that relies on moments. The concept is to study the difference between the data moments and the model moments using a matrix norm, such as the Frobenius norm. We show how to use this new metric for model comparison and then for regularization. It is common to draw samples from the fitted distribution when evaluating latent variable models and we show that our proposed metric is faster to compute and has a smaller variance that this alternative. We conclude this article with a proof of concept of both applications and we discuss future work." @default.
- W4376982954 created "2023-05-18" @default.
- W4376982954 creator A5011407083 @default.
- W4376982954 date "2023-06-12" @default.
- W4376982954 modified "2023-10-18" @default.
- W4376982954 title "A moment-matching metric for latent variable generative models" @default.
- W4376982954 cites W2000661457 @default.
- W4376982954 cites W2015887370 @default.
- W4376982954 cites W2058815839 @default.
- W4376982954 cites W2112796928 @default.
- W4376982954 cites W2125027820 @default.
- W4376982954 cites W2129286663 @default.
- W4376982954 cites W2142635246 @default.
- W4376982954 cites W2157076315 @default.
- W4376982954 cites W2296719434 @default.
- W4376982954 cites W2472097883 @default.
- W4376982954 cites W2787894218 @default.
- W4376982954 cites W2903782687 @default.
- W4376982954 cites W3099878876 @default.
- W4376982954 cites W4205370808 @default.
- W4376982954 cites W4236362309 @default.
- W4376982954 cites W4294541781 @default.
- W4376982954 doi "https://doi.org/10.1007/s10994-023-06340-x" @default.
- W4376982954 hasPublicationYear "2023" @default.
- W4376982954 type Work @default.
- W4376982954 citedByCount "0" @default.
- W4376982954 crossrefType "journal-article" @default.
- W4376982954 hasAuthorship W4376982954A5011407083 @default.
- W4376982954 hasBestOaLocation W43769829541 @default.
- W4376982954 hasConcept C11413529 @default.
- W4376982954 hasConcept C119857082 @default.
- W4376982954 hasConcept C121332964 @default.
- W4376982954 hasConcept C154945302 @default.
- W4376982954 hasConcept C158693339 @default.
- W4376982954 hasConcept C162324750 @default.
- W4376982954 hasConcept C163716315 @default.
- W4376982954 hasConcept C176217482 @default.
- W4376982954 hasConcept C179254644 @default.
- W4376982954 hasConcept C21547014 @default.
- W4376982954 hasConcept C2776135515 @default.
- W4376982954 hasConcept C28826006 @default.
- W4376982954 hasConcept C33923547 @default.
- W4376982954 hasConcept C41008148 @default.
- W4376982954 hasConcept C51167844 @default.
- W4376982954 hasConcept C62520636 @default.
- W4376982954 hasConcept C65965080 @default.
- W4376982954 hasConcept C74650414 @default.
- W4376982954 hasConcept C92207270 @default.
- W4376982954 hasConceptScore W4376982954C11413529 @default.
- W4376982954 hasConceptScore W4376982954C119857082 @default.
- W4376982954 hasConceptScore W4376982954C121332964 @default.
- W4376982954 hasConceptScore W4376982954C154945302 @default.
- W4376982954 hasConceptScore W4376982954C158693339 @default.
- W4376982954 hasConceptScore W4376982954C162324750 @default.
- W4376982954 hasConceptScore W4376982954C163716315 @default.
- W4376982954 hasConceptScore W4376982954C176217482 @default.
- W4376982954 hasConceptScore W4376982954C179254644 @default.
- W4376982954 hasConceptScore W4376982954C21547014 @default.
- W4376982954 hasConceptScore W4376982954C2776135515 @default.
- W4376982954 hasConceptScore W4376982954C28826006 @default.
- W4376982954 hasConceptScore W4376982954C33923547 @default.
- W4376982954 hasConceptScore W4376982954C41008148 @default.
- W4376982954 hasConceptScore W4376982954C51167844 @default.
- W4376982954 hasConceptScore W4376982954C62520636 @default.
- W4376982954 hasConceptScore W4376982954C65965080 @default.
- W4376982954 hasConceptScore W4376982954C74650414 @default.
- W4376982954 hasConceptScore W4376982954C92207270 @default.
- W4376982954 hasIssue "10" @default.
- W4376982954 hasLocation W43769829541 @default.
- W4376982954 hasLocation W43769829542 @default.
- W4376982954 hasLocation W43769829543 @default.
- W4376982954 hasOpenAccess W4376982954 @default.
- W4376982954 hasPrimaryLocation W43769829541 @default.
- W4376982954 hasRelatedWork W2003332620 @default.
- W4376982954 hasRelatedWork W2041580163 @default.
- W4376982954 hasRelatedWork W2085068179 @default.
- W4376982954 hasRelatedWork W2549396809 @default.
- W4376982954 hasRelatedWork W2786010873 @default.
- W4376982954 hasRelatedWork W2963529210 @default.
- W4376982954 hasRelatedWork W3041775337 @default.
- W4376982954 hasRelatedWork W3137809915 @default.
- W4376982954 hasRelatedWork W4211155169 @default.
- W4376982954 hasRelatedWork W4293659211 @default.
- W4376982954 hasVolume "112" @default.
- W4376982954 isParatext "false" @default.
- W4376982954 isRetracted "false" @default.
- W4376982954 workType "article" @default.