Matches in SemOpenAlex for { <https://semopenalex.org/work/W4283314575> ?p ?o ?g. }
Showing items 1 to 55 of
55
with 100 items per page.
- W4283314575 abstract "Algorithmic Gaussianization is a phenomenon that can arise when using randomized sketching or sampling methods to produce smaller representations of large datasets: For certain tasks, these sketched representations have been observed to exhibit many robust performance characteristics that are known to occur when a data sample comes from a sub-gaussian random design, which is a powerful statistical model of data distributions. However, this phenomenon has only been studied for specific tasks and metrics, or by relying on computationally expensive methods. We address this by providing an algorithmic framework for gaussianizing data distributions via averaging, proving that it is possible to efficiently construct data sketches that are nearly indistinguishable (in terms of total variation distance) from sub-gaussian random designs. In particular, relying on a recently introduced sketching technique called Leverage Score Sparsified (LESS) embeddings, we show that one can construct an $ntimes d$ sketch of an $Ntimes d$ matrix $A$, where $nll N$, that is nearly indistinguishable from a sub-gaussian design, in time $O(text{nnz}(A)log N + nd^2)$, where $text{nnz}(A)$ is the number of non-zero entries in $A$. As a consequence, strong statistical guarantees and precise asymptotics available for the estimators produced from sub-gaussian designs (e.g., for least squares and Lasso regression, covariance estimation, low-rank approximation, etc.) can be straightforwardly adapted to our sketching framework. We illustrate this with a new approximation guarantee for sketched least squares, among other examples." @default.
- W4283314575 created "2022-06-24" @default.
- W4283314575 creator A5013422880 @default.
- W4283314575 date "2022-06-21" @default.
- W4283314575 modified "2023-09-26" @default.
- W4283314575 title "Algorithmic Gaussianization through Sketching: Converting Data into Sub-gaussian Random Designs" @default.
- W4283314575 doi "https://doi.org/10.48550/arxiv.2206.10291" @default.
- W4283314575 hasPublicationYear "2022" @default.
- W4283314575 type Work @default.
- W4283314575 citedByCount "0" @default.
- W4283314575 crossrefType "posted-content" @default.
- W4283314575 hasAuthorship W4283314575A5013422880 @default.
- W4283314575 hasBestOaLocation W42833145751 @default.
- W4283314575 hasConcept C105795698 @default.
- W4283314575 hasConcept C11413529 @default.
- W4283314575 hasConcept C121332964 @default.
- W4283314575 hasConcept C153083717 @default.
- W4283314575 hasConcept C154945302 @default.
- W4283314575 hasConcept C163716315 @default.
- W4283314575 hasConcept C178650346 @default.
- W4283314575 hasConcept C185429906 @default.
- W4283314575 hasConcept C199360897 @default.
- W4283314575 hasConcept C2780801425 @default.
- W4283314575 hasConcept C33923547 @default.
- W4283314575 hasConcept C41008148 @default.
- W4283314575 hasConcept C62520636 @default.
- W4283314575 hasConceptScore W4283314575C105795698 @default.
- W4283314575 hasConceptScore W4283314575C11413529 @default.
- W4283314575 hasConceptScore W4283314575C121332964 @default.
- W4283314575 hasConceptScore W4283314575C153083717 @default.
- W4283314575 hasConceptScore W4283314575C154945302 @default.
- W4283314575 hasConceptScore W4283314575C163716315 @default.
- W4283314575 hasConceptScore W4283314575C178650346 @default.
- W4283314575 hasConceptScore W4283314575C185429906 @default.
- W4283314575 hasConceptScore W4283314575C199360897 @default.
- W4283314575 hasConceptScore W4283314575C2780801425 @default.
- W4283314575 hasConceptScore W4283314575C33923547 @default.
- W4283314575 hasConceptScore W4283314575C41008148 @default.
- W4283314575 hasConceptScore W4283314575C62520636 @default.
- W4283314575 hasLocation W42833145751 @default.
- W4283314575 hasOpenAccess W4283314575 @default.
- W4283314575 hasPrimaryLocation W42833145751 @default.
- W4283314575 hasRelatedWork W1969324738 @default.
- W4283314575 hasRelatedWork W1988224349 @default.
- W4283314575 hasRelatedWork W2024105718 @default.
- W4283314575 hasRelatedWork W2351571780 @default.
- W4283314575 hasRelatedWork W2353495529 @default.
- W4283314575 hasRelatedWork W3131901029 @default.
- W4283314575 hasRelatedWork W4296285654 @default.
- W4283314575 hasRelatedWork W4302764549 @default.
- W4283314575 hasRelatedWork W4361233137 @default.
- W4283314575 hasRelatedWork W4364354367 @default.
- W4283314575 isParatext "false" @default.
- W4283314575 isRetracted "false" @default.
- W4283314575 workType "article" @default.