Matches in SemOpenAlex for { <https://semopenalex.org/work/W3124964988> ?p ?o ?g. }
- W3124964988 abstract "Convolutional image classifiers can achieve high predictive accuracy, but quantifying their uncertainty remains an unresolved challenge, hindering their deployment in consequential settings. Existing uncertainty quantification techniques, such as Platt scaling, attempt to calibrate the network’s probability estimates, but they do not have formal guarantees. We present an algorithm that modifies any classifier to output a predictive set containing the true label with a user-specified probability, such as 90%. The algorithm is simple and fast like Platt scaling, but provides a formal finite-sample coverage guarantee for every model and dataset. Our method modifies an existing conformal prediction algorithm to give more stable predictive sets by regularizing the small scores of unlikely classes after Platt scaling. In experiments on both Imagenet and Imagenet-V2 with ResNet-152 and other classifiers, our scheme outperforms existing approaches, achieving coverage with sets that are often factors of 5 to 10 smaller." @default.
- W3124964988 created "2021-02-01" @default.
- W3124964988 creator A5001594573 @default.
- W3124964988 creator A5019661690 @default.
- W3124964988 creator A5045046961 @default.
- W3124964988 creator A5049812527 @default.
- W3124964988 date "2021-05-03" @default.
- W3124964988 modified "2023-09-24" @default.
- W3124964988 title "Uncertainty Sets for Image Classifiers using Conformal Prediction" @default.
- W3124964988 cites W1488642921 @default.
- W3124964988 cites W1499689786 @default.
- W3124964988 cites W1553101044 @default.
- W3124964988 cites W1567512734 @default.
- W3124964988 cites W1618905105 @default.
- W3124964988 cites W1934410531 @default.
- W3124964988 cites W1990621008 @default.
- W3124964988 cites W2044097773 @default.
- W3124964988 cites W2117670920 @default.
- W3124964988 cites W2158471180 @default.
- W3124964988 cites W2577719032 @default.
- W3124964988 cites W2607941059 @default.
- W3124964988 cites W2806471870 @default.
- W3124964988 cites W2810469995 @default.
- W3124964988 cites W2962995943 @default.
- W3124964988 cites W2963238274 @default.
- W3124964988 cites W2963441669 @default.
- W3124964988 cites W2964330179 @default.
- W3124964988 cites W2969338701 @default.
- W3124964988 cites W2970311848 @default.
- W3124964988 cites W2970971581 @default.
- W3124964988 cites W2971169014 @default.
- W3124964988 cites W2996766289 @default.
- W3124964988 cites W3021498541 @default.
- W3124964988 cites W3027782004 @default.
- W3124964988 cites W3034127978 @default.
- W3124964988 cites W3034151283 @default.
- W3124964988 cites W3041560869 @default.
- W3124964988 hasPublicationYear "2021" @default.
- W3124964988 type Work @default.
- W3124964988 sameAs 3124964988 @default.
- W3124964988 citedByCount "7" @default.
- W3124964988 countsByYear W31249649882020 @default.
- W3124964988 countsByYear W31249649882021 @default.
- W3124964988 crossrefType "proceedings-article" @default.
- W3124964988 hasAuthorship W3124964988A5001594573 @default.
- W3124964988 hasAuthorship W3124964988A5019661690 @default.
- W3124964988 hasAuthorship W3124964988A5045046961 @default.
- W3124964988 hasAuthorship W3124964988A5049812527 @default.
- W3124964988 hasConcept C11413529 @default.
- W3124964988 hasConcept C115961682 @default.
- W3124964988 hasConcept C119857082 @default.
- W3124964988 hasConcept C124101348 @default.
- W3124964988 hasConcept C134306372 @default.
- W3124964988 hasConcept C153180895 @default.
- W3124964988 hasConcept C154945302 @default.
- W3124964988 hasConcept C177264268 @default.
- W3124964988 hasConcept C199360897 @default.
- W3124964988 hasConcept C2524010 @default.
- W3124964988 hasConcept C33923547 @default.
- W3124964988 hasConcept C41008148 @default.
- W3124964988 hasConcept C75294576 @default.
- W3124964988 hasConcept C95623464 @default.
- W3124964988 hasConcept C98214594 @default.
- W3124964988 hasConcept C99844830 @default.
- W3124964988 hasConceptScore W3124964988C11413529 @default.
- W3124964988 hasConceptScore W3124964988C115961682 @default.
- W3124964988 hasConceptScore W3124964988C119857082 @default.
- W3124964988 hasConceptScore W3124964988C124101348 @default.
- W3124964988 hasConceptScore W3124964988C134306372 @default.
- W3124964988 hasConceptScore W3124964988C153180895 @default.
- W3124964988 hasConceptScore W3124964988C154945302 @default.
- W3124964988 hasConceptScore W3124964988C177264268 @default.
- W3124964988 hasConceptScore W3124964988C199360897 @default.
- W3124964988 hasConceptScore W3124964988C2524010 @default.
- W3124964988 hasConceptScore W3124964988C33923547 @default.
- W3124964988 hasConceptScore W3124964988C41008148 @default.
- W3124964988 hasConceptScore W3124964988C75294576 @default.
- W3124964988 hasConceptScore W3124964988C95623464 @default.
- W3124964988 hasConceptScore W3124964988C98214594 @default.
- W3124964988 hasConceptScore W3124964988C99844830 @default.
- W3124964988 hasLocation W31249649881 @default.
- W3124964988 hasOpenAccess W3124964988 @default.
- W3124964988 hasPrimaryLocation W31249649881 @default.
- W3124964988 hasRelatedWork W1553101044 @default.
- W3124964988 hasRelatedWork W2056671638 @default.
- W3124964988 hasRelatedWork W2146156326 @default.
- W3124964988 hasRelatedWork W2514278201 @default.
- W3124964988 hasRelatedWork W2516932512 @default.
- W3124964988 hasRelatedWork W2519997192 @default.
- W3124964988 hasRelatedWork W2747669017 @default.
- W3124964988 hasRelatedWork W2758425594 @default.
- W3124964988 hasRelatedWork W2782944630 @default.
- W3124964988 hasRelatedWork W2950832426 @default.
- W3124964988 hasRelatedWork W2963058377 @default.
- W3124964988 hasRelatedWork W2970311848 @default.
- W3124964988 hasRelatedWork W2971169014 @default.
- W3124964988 hasRelatedWork W3083184539 @default.
- W3124964988 hasRelatedWork W3084456568 @default.
- W3124964988 hasRelatedWork W3089746993 @default.
- W3124964988 hasRelatedWork W3097199945 @default.