Matches in SemOpenAlex for { <https://semopenalex.org/work/W4221154731> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4221154731 abstract "We give the first sample complexity characterizations for outcome indistinguishability, a theoretical framework of machine learning recently introduced by Dwork, Kim, Reingold, Rothblum, and Yona (STOC 2021). In outcome indistinguishability, the goal of the learner is to output a predictor that cannot be distinguished from the target predictor by a class $D$ of distinguishers examining the outcomes generated according to the predictors' predictions. In the distribution-specific and realizable setting where the learner is given the data distribution together with a predictor class $P$ containing the target predictor, we show that the sample complexity of outcome indistinguishability is characterized by the metric entropy of $P$ w.r.t. the dual Minkowski norm defined by $D$, and equivalently by the metric entropy of $D$ w.r.t. the dual Minkowski norm defined by $P$. This equivalence makes an intriguing connection to the long-standing metric entropy duality conjecture in convex geometry. Our sample complexity characterization implies a variant of metric entropy duality, which we show is nearly tight. In the distribution-free setting, we focus on the case considered by Dwork et al. where $P$ contains all possible predictors, hence the sample complexity only depends on $D$. In this setting, we show that the sample complexity of outcome indistinguishability is characterized by the fat-shattering dimension of $D$. We also show a strong sample complexity separation between realizable and agnostic outcome indistinguishability in both the distribution-free and the distribution-specific settings. This is in contrast to distribution-free (resp. distribution-specific) PAC learning where the sample complexity in both the realizable and the agnostic settings can be characterized by the VC dimension (resp. metric entropy)." @default.
- W4221154731 created "2022-04-03" @default.
- W4221154731 creator A5006847182 @default.
- W4221154731 creator A5076540826 @default.
- W4221154731 creator A5086367709 @default.
- W4221154731 date "2022-03-09" @default.
- W4221154731 modified "2023-10-18" @default.
- W4221154731 title "Metric Entropy Duality and the Sample Complexity of Outcome Indistinguishability" @default.
- W4221154731 doi "https://doi.org/10.48550/arxiv.2203.04536" @default.
- W4221154731 hasPublicationYear "2022" @default.
- W4221154731 type Work @default.
- W4221154731 citedByCount "0" @default.
- W4221154731 crossrefType "posted-content" @default.
- W4221154731 hasAuthorship W4221154731A5006847182 @default.
- W4221154731 hasAuthorship W4221154731A5076540826 @default.
- W4221154731 hasAuthorship W4221154731A5086367709 @default.
- W4221154731 hasBestOaLocation W42211547311 @default.
- W4221154731 hasConcept C106301342 @default.
- W4221154731 hasConcept C114614502 @default.
- W4221154731 hasConcept C118615104 @default.
- W4221154731 hasConcept C121332964 @default.
- W4221154731 hasConcept C144237770 @default.
- W4221154731 hasConcept C148220186 @default.
- W4221154731 hasConcept C154945302 @default.
- W4221154731 hasConcept C162324750 @default.
- W4221154731 hasConcept C176217482 @default.
- W4221154731 hasConcept C21547014 @default.
- W4221154731 hasConcept C2524010 @default.
- W4221154731 hasConcept C2778023678 @default.
- W4221154731 hasConcept C2778445095 @default.
- W4221154731 hasConcept C2780990831 @default.
- W4221154731 hasConcept C33923547 @default.
- W4221154731 hasConcept C41008148 @default.
- W4221154731 hasConcept C62520636 @default.
- W4221154731 hasConcept C79464548 @default.
- W4221154731 hasConceptScore W4221154731C106301342 @default.
- W4221154731 hasConceptScore W4221154731C114614502 @default.
- W4221154731 hasConceptScore W4221154731C118615104 @default.
- W4221154731 hasConceptScore W4221154731C121332964 @default.
- W4221154731 hasConceptScore W4221154731C144237770 @default.
- W4221154731 hasConceptScore W4221154731C148220186 @default.
- W4221154731 hasConceptScore W4221154731C154945302 @default.
- W4221154731 hasConceptScore W4221154731C162324750 @default.
- W4221154731 hasConceptScore W4221154731C176217482 @default.
- W4221154731 hasConceptScore W4221154731C21547014 @default.
- W4221154731 hasConceptScore W4221154731C2524010 @default.
- W4221154731 hasConceptScore W4221154731C2778023678 @default.
- W4221154731 hasConceptScore W4221154731C2778445095 @default.
- W4221154731 hasConceptScore W4221154731C2780990831 @default.
- W4221154731 hasConceptScore W4221154731C33923547 @default.
- W4221154731 hasConceptScore W4221154731C41008148 @default.
- W4221154731 hasConceptScore W4221154731C62520636 @default.
- W4221154731 hasConceptScore W4221154731C79464548 @default.
- W4221154731 hasLocation W42211547311 @default.
- W4221154731 hasOpenAccess W4221154731 @default.
- W4221154731 hasPrimaryLocation W42211547311 @default.
- W4221154731 hasRelatedWork W2013467167 @default.
- W4221154731 hasRelatedWork W2029981401 @default.
- W4221154731 hasRelatedWork W2740334431 @default.
- W4221154731 hasRelatedWork W2951982256 @default.
- W4221154731 hasRelatedWork W2962701216 @default.
- W4221154731 hasRelatedWork W2963229517 @default.
- W4221154731 hasRelatedWork W3123287700 @default.
- W4221154731 hasRelatedWork W4212808623 @default.
- W4221154731 hasRelatedWork W4294890818 @default.
- W4221154731 hasRelatedWork W4301344670 @default.
- W4221154731 isParatext "false" @default.
- W4221154731 isRetracted "false" @default.
- W4221154731 workType "article" @default.