FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2942302803> ?p ?o ?g. }

• W2942302803 endingPage "902" @default.
• W2942302803 startingPage "897" @default.
• W2942302803 abstract "Minimizing a convex risk function is the main step in many basic learning algorithms. We study protocols for convex optimization which provably leak very little about the individual data points that constitute the loss function. Specifically, we consider differentially private algorithms that operate in the local model, where each data record is stored on a separate user device and randomization is performed locally by those devices. We give new protocols for emph{noninteractive} LDP convex optimization---i.e., protocols that require only a single randomized report from each user to an untrusted aggregator. We study our algorithms' performance with respect to expected loss---either over the data set at hand (empirical risk) or a larger population from which our data set is assumed to be drawn. Our error bounds depend on the form of individuals' contribution to the expected loss. For the case of emph{generalized linear losses} (such as hinge and logistic losses), we give an LDP algorithm whose sample complexity is only linear in the dimensionality $p$ and quasipolynomial in other terms (the privacy parameters $epsilon$ and $delta$, and the desired excess risk $alpha$). This is the first algorithm for nonsmooth losses with sub-exponential dependence on $p$. For the Euclidean median problem, where the loss is given by the Euclidean distance to a given data point, we give a protocol whose sample complexity grows quasipolynomially in $p$. This is the first protocol with sub-exponential dependence on $p$ for a loss that is not a generalized linear loss . Our result for the hinge loss is based on a technique, dubbed polynomial of inner product approximation, which may be applicable to other problems. Our results for generalized linear losses and the Euclidean median are based on new reductions to the case of hinge loss." @default.
• W2942302803 created "2019-05-03" @default.
• W2942302803 creator A5006945148 @default.
• W2942302803 creator A5052304130 @default.
• W2942302803 creator A5079131956 @default.
• W2942302803 date "2019-03-10" @default.
• W2942302803 modified "2023-09-22" @default.
• W2942302803 title "Noninteractive Locally Private Learning of Linear Models via Polynomial Approximations." @default.
• W2942302803 hasPublicationYear "2019" @default.
• W2942302803 type Work @default.
• W2942302803 sameAs 2942302803 @default.
• W2942302803 citedByCount "9" @default.
• W2942302803 countsByYear W29423028032019 @default.
• W2942302803 countsByYear W29423028032020 @default.
• W2942302803 countsByYear W29423028032021 @default.
• W2942302803 crossrefType "proceedings-article" @default.
• W2942302803 hasAuthorship W2942302803A5006945148 @default.
• W2942302803 hasAuthorship W2942302803A5052304130 @default.
• W2942302803 hasAuthorship W2942302803A5079131956 @default.
• W2942302803 hasConcept C11413529 @default.
• W2942302803 hasConcept C118615104 @default.
• W2942302803 hasConcept C134306372 @default.
• W2942302803 hasConcept C144024400 @default.
• W2942302803 hasConcept C149923435 @default.
• W2942302803 hasConcept C2908647359 @default.
• W2942302803 hasConcept C33923547 @default.
• W2942302803 hasConcept C34388435 @default.
• W2942302803 hasConcept C41008148 @default.
• W2942302803 hasConcept C90119067 @default.
• W2942302803 hasConceptScore W2942302803C11413529 @default.
• W2942302803 hasConceptScore W2942302803C118615104 @default.
• W2942302803 hasConceptScore W2942302803C134306372 @default.
• W2942302803 hasConceptScore W2942302803C144024400 @default.
• W2942302803 hasConceptScore W2942302803C149923435 @default.
• W2942302803 hasConceptScore W2942302803C2908647359 @default.
• W2942302803 hasConceptScore W2942302803C33923547 @default.
• W2942302803 hasConceptScore W2942302803C34388435 @default.
• W2942302803 hasConceptScore W2942302803C41008148 @default.
• W2942302803 hasConceptScore W2942302803C90119067 @default.
• W2942302803 hasLocation W29423028031 @default.
• W2942302803 hasOpenAccess W2942302803 @default.
• W2942302803 hasPrimaryLocation W29423028031 @default.
• W2942302803 hasRelatedWork W1873763122 @default.
• W2942302803 hasRelatedWork W1992926795 @default.
• W2942302803 hasRelatedWork W2027595342 @default.
• W2942302803 hasRelatedWork W2053801139 @default.
• W2942302803 hasRelatedWork W2112380340 @default.
• W2942302803 hasRelatedWork W2119874464 @default.
• W2942302803 hasRelatedWork W2129658066 @default.
• W2942302803 hasRelatedWork W2182267394 @default.
• W2942302803 hasRelatedWork W2184139426 @default.
• W2942302803 hasRelatedWork W2245160765 @default.
• W2942302803 hasRelatedWork W2470101972 @default.
• W2942302803 hasRelatedWork W2473418344 @default.
• W2942302803 hasRelatedWork W2652948231 @default.
• W2942302803 hasRelatedWork W2751484150 @default.
• W2942302803 hasRelatedWork W2804827666 @default.
• W2942302803 hasRelatedWork W2891848848 @default.
• W2942302803 hasRelatedWork W2904039766 @default.
• W2942302803 hasRelatedWork W2963695762 @default.
• W2942302803 hasRelatedWork W2975069372 @default.
• W2942302803 hasRelatedWork W3102407811 @default.
• W2942302803 isParatext "false" @default.
• W2942302803 isRetracted "false" @default.
• W2942302803 magId "2942302803" @default.
• W2942302803 workType "article" @default.