Matches in SemOpenAlex for { <https://semopenalex.org/work/W3101595931> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W3101595931 endingPage "4234" @default.
- W3101595931 startingPage "4223" @default.
- W3101595931 abstract "Commonly used classification algorithms in machine learning, such as support vector machines, minimize a convex surrogate loss on training examples. In practice, these algorithms are surprisingly robust to errors in the training data. In this work, we identify a set of conditions on the data under which such surrogate loss minimization algorithms provably learn the correct classifier. This allows us to establish, in a unified framework, the robustness of these algorithms under various models on data as well as error. In particular, we show that if the data is linearly classifiable with a slightly non-trivial margin (i.e. a margin at least $C/sqrt{d}$ for $d$-dimensional unit vectors), and the class-conditional distributions are near isotropic and logconcave, then surrogate loss minimization has negligible error on the uncorrupted data even when a constant fraction of examples are adversarially mislabeled." @default.
- W3101595931 created "2020-11-23" @default.
- W3101595931 creator A5080065269 @default.
- W3101595931 date "2020-01-01" @default.
- W3101595931 modified "2023-10-18" @default.
- W3101595931 title "On the Error Resistance of Hinge-Loss Minimization" @default.
- W3101595931 hasPublicationYear "2020" @default.
- W3101595931 type Work @default.
- W3101595931 sameAs 3101595931 @default.
- W3101595931 citedByCount "1" @default.
- W3101595931 countsByYear W31015959312020 @default.
- W3101595931 crossrefType "proceedings-article" @default.
- W3101595931 hasAuthorship W3101595931A5080065269 @default.
- W3101595931 hasConcept C104317684 @default.
- W3101595931 hasConcept C107321475 @default.
- W3101595931 hasConcept C11413529 @default.
- W3101595931 hasConcept C119857082 @default.
- W3101595931 hasConcept C12267149 @default.
- W3101595931 hasConcept C126255220 @default.
- W3101595931 hasConcept C147764199 @default.
- W3101595931 hasConcept C154945302 @default.
- W3101595931 hasConcept C185592680 @default.
- W3101595931 hasConcept C33923547 @default.
- W3101595931 hasConcept C39891107 @default.
- W3101595931 hasConcept C41008148 @default.
- W3101595931 hasConcept C51632099 @default.
- W3101595931 hasConcept C55493867 @default.
- W3101595931 hasConcept C63479239 @default.
- W3101595931 hasConcept C774472 @default.
- W3101595931 hasConcept C95623464 @default.
- W3101595931 hasConceptScore W3101595931C104317684 @default.
- W3101595931 hasConceptScore W3101595931C107321475 @default.
- W3101595931 hasConceptScore W3101595931C11413529 @default.
- W3101595931 hasConceptScore W3101595931C119857082 @default.
- W3101595931 hasConceptScore W3101595931C12267149 @default.
- W3101595931 hasConceptScore W3101595931C126255220 @default.
- W3101595931 hasConceptScore W3101595931C147764199 @default.
- W3101595931 hasConceptScore W3101595931C154945302 @default.
- W3101595931 hasConceptScore W3101595931C185592680 @default.
- W3101595931 hasConceptScore W3101595931C33923547 @default.
- W3101595931 hasConceptScore W3101595931C39891107 @default.
- W3101595931 hasConceptScore W3101595931C41008148 @default.
- W3101595931 hasConceptScore W3101595931C51632099 @default.
- W3101595931 hasConceptScore W3101595931C55493867 @default.
- W3101595931 hasConceptScore W3101595931C63479239 @default.
- W3101595931 hasConceptScore W3101595931C774472 @default.
- W3101595931 hasConceptScore W3101595931C95623464 @default.
- W3101595931 hasLocation W31015959311 @default.
- W3101595931 hasOpenAccess W3101595931 @default.
- W3101595931 hasPrimaryLocation W31015959311 @default.
- W3101595931 hasRelatedWork W1898805148 @default.
- W3101595931 hasRelatedWork W2099210013 @default.
- W3101595931 hasRelatedWork W2107961375 @default.
- W3101595931 hasRelatedWork W2247593035 @default.
- W3101595931 hasRelatedWork W2472226770 @default.
- W3101595931 hasRelatedWork W2766159815 @default.
- W3101595931 hasRelatedWork W2897313170 @default.
- W3101595931 hasRelatedWork W2963147719 @default.
- W3101595931 hasRelatedWork W2963424904 @default.
- W3101595931 hasRelatedWork W2963705276 @default.
- W3101595931 hasRelatedWork W2964008913 @default.
- W3101595931 hasRelatedWork W2964933009 @default.
- W3101595931 hasRelatedWork W2986048261 @default.
- W3101595931 hasRelatedWork W3003022452 @default.
- W3101595931 hasRelatedWork W3031535528 @default.
- W3101595931 hasRelatedWork W3040534566 @default.
- W3101595931 hasRelatedWork W3107593056 @default.
- W3101595931 hasRelatedWork W3160748929 @default.
- W3101595931 hasRelatedWork W3213495149 @default.
- W3101595931 hasRelatedWork W329869691 @default.
- W3101595931 hasVolume "33" @default.
- W3101595931 isParatext "false" @default.
- W3101595931 isRetracted "false" @default.
- W3101595931 magId "3101595931" @default.
- W3101595931 workType "article" @default.