FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3129100675> ?p ?o ?g. }

• W3129100675 abstract "The label noise transition matrix, characterizing the probabilities of a training instance being wrongly annotated, is crucial to designing popular solutions to learning with noisy labels. Existing works heavily rely on finding or their approximates, defined as instances belonging to a particular class almost surely. Nonetheless, finding anchor points remains a non-trivial task, and the estimation accuracy is also often throttled by the number of available anchor points. In this paper, we propose an alternative option to the above task. Our main contribution is the discovery of an efficient estimation procedure based on a clusterability condition. We prove that with clusterable representations of features, using up to third-order consensuses of noisy labels among neighbor representations is sufficient to estimate a unique transition matrix. Compared with methods using anchor points, our approach uses substantially more instances and benefits from a much better sample complexity. We demonstrate the estimation accuracy and advantages of our estimates using both synthetic noisy labels (on CIFAR-10/100) and real human-level noisy labels (on Clothing1M and our self-collected human-annotated CIFAR-10). Our code and human-level noisy CIFAR-10 labels are available at this https URL." @default.
• W3129100675 created "2021-02-15" @default.
• W3129100675 creator A5023363049 @default.
• W3129100675 creator A5025441627 @default.
• W3129100675 creator A5054105523 @default.
• W3129100675 date "2021-02-10" @default.
• W3129100675 modified "2023-09-27" @default.
• W3129100675 title "Clusterability as an Alternative to Anchor Points When Learning with Noisy Labels" @default.
• W3129100675 cites W1514928307 @default.
• W3129100675 cites W1921293667 @default.
• W3129100675 cites W1989135160 @default.
• W3129100675 cites W2103012681 @default.
• W3129100675 cites W2113290770 @default.
• W3129100675 cites W2129345386 @default.
• W3129100675 cites W2163922914 @default.
• W3129100675 cites W2194775991 @default.
• W3129100675 cites W2611425310 @default.
• W3129100675 cites W2708237239 @default.
• W3129100675 cites W2810089099 @default.
• W3129100675 cites W2886811283 @default.
• W3129100675 cites W2891461240 @default.
• W3129100675 cites W2913939497 @default.
• W3129100675 cites W2963081269 @default.
• W3129100675 cites W2963697299 @default.
• W3129100675 cites W2963735582 @default.
• W3129100675 cites W2963759070 @default.
• W3129100675 cites W2964292098 @default.
• W3129100675 cites W2970038028 @default.
• W3129100675 cites W2970308742 @default.
• W3129100675 cites W2971193222 @default.
• W3129100675 cites W2981873476 @default.
• W3129100675 cites W2990019157 @default.
• W3129100675 cites W2990500698 @default.
• W3129100675 cites W3005871659 @default.
• W3129100675 cites W303151045 @default.
• W3129100675 cites W3034185248 @default.
• W3129100675 cites W3034432520 @default.
• W3129100675 cites W3035261884 @default.
• W3129100675 cites W3035314656 @default.
• W3129100675 cites W3035695020 @default.
• W3129100675 cites W3035753488 @default.
• W3129100675 cites W3041687329 @default.
• W3129100675 cites W3102643535 @default.
• W3129100675 cites W3103846556 @default.
• W3129100675 cites W3104255659 @default.
• W3129100675 cites W3110483780 @default.
• W3129100675 cites W3114070207 @default.
• W3129100675 cites W3118608800 @default.
• W3129100675 cites W3123795136 @default.
• W3129100675 cites W3127098609 @default.
• W3129100675 cites W3128198050 @default.
• W3129100675 cites W3129603874 @default.
• W3129100675 cites W3134631405 @default.
• W3129100675 cites W3156669901 @default.
• W3129100675 cites W3166626599 @default.
• W3129100675 cites W3170414933 @default.
• W3129100675 cites W3171240668 @default.
• W3129100675 cites W3122805849 @default.
• W3129100675 hasPublicationYear "2021" @default.
• W3129100675 type Work @default.
• W3129100675 sameAs 3129100675 @default.
• W3129100675 citedByCount "11" @default.
• W3129100675 countsByYear W31291006752020 @default.
• W3129100675 countsByYear W31291006752021 @default.
• W3129100675 crossrefType "posted-content" @default.
• W3129100675 hasAuthorship W3129100675A5023363049 @default.
• W3129100675 hasAuthorship W3129100675A5025441627 @default.
• W3129100675 hasAuthorship W3129100675A5054105523 @default.
• W3129100675 hasConcept C106487976 @default.
• W3129100675 hasConcept C11413529 @default.
• W3129100675 hasConcept C115961682 @default.
• W3129100675 hasConcept C119857082 @default.
• W3129100675 hasConcept C153180895 @default.
• W3129100675 hasConcept C154945302 @default.
• W3129100675 hasConcept C159985019 @default.
• W3129100675 hasConcept C162324750 @default.
• W3129100675 hasConcept C177264268 @default.
• W3129100675 hasConcept C187736073 @default.
• W3129100675 hasConcept C192562407 @default.
• W3129100675 hasConcept C199360897 @default.
• W3129100675 hasConcept C2776760102 @default.
• W3129100675 hasConcept C2777212361 @default.
• W3129100675 hasConcept C2780451532 @default.
• W3129100675 hasConcept C41008148 @default.
• W3129100675 hasConcept C49555168 @default.
• W3129100675 hasConcept C98763669 @default.
• W3129100675 hasConcept C99498987 @default.
• W3129100675 hasConceptScore W3129100675C106487976 @default.
• W3129100675 hasConceptScore W3129100675C11413529 @default.
• W3129100675 hasConceptScore W3129100675C115961682 @default.
• W3129100675 hasConceptScore W3129100675C119857082 @default.
• W3129100675 hasConceptScore W3129100675C153180895 @default.
• W3129100675 hasConceptScore W3129100675C154945302 @default.
• W3129100675 hasConceptScore W3129100675C159985019 @default.
• W3129100675 hasConceptScore W3129100675C162324750 @default.
• W3129100675 hasConceptScore W3129100675C177264268 @default.
• W3129100675 hasConceptScore W3129100675C187736073 @default.
• W3129100675 hasConceptScore W3129100675C192562407 @default.
• W3129100675 hasConceptScore W3129100675C199360897 @default.
• W3129100675 hasConceptScore W3129100675C2776760102 @default.

• next