Matches in SemOpenAlex for { <https://semopenalex.org/work/W3103183085> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W3103183085 endingPage "19387" @default.
- W3103183085 startingPage "19377" @default.
- W3103183085 abstract "We propose a new randomized algorithm for solving L2-regularized least-squares problems based on sketching. We consider two of the most popular random embeddings, namely, Gaussian embeddings and the Subsampled Randomized Hadamard Transform (SRHT). While current randomized solvers for least-squares optimization prescribe an embedding dimension at least greater than the data dimension, we show that the embedding dimension can be reduced to the effective dimension of the optimization problem, and still preserve high-probability convergence guarantees. In this regard, we derive sharp matrix deviation inequalities over ellipsoids for both Gaussian and SRHT embeddings. Specifically, we improve on the constant of a classical Gaussian concentration bound whereas, for SRHT embeddings, our deviation inequality involves a novel technical approach. Leveraging these bounds, we are able to design a practical and adaptive algorithm which does not require to know the effective dimension beforehand. Our method starts with an initial embedding dimension equal to 1 and, over iterations, increases the embedding dimension up to the effective one at most. Hence, our algorithm improves the state-of-the-art computational complexity for solving regularized least-squares problems. Further, we show numerically that it outperforms standard iterative solvers such as the conjugate gradient method and its pre-conditioned version on several standard machine learning datasets." @default.
- W3103183085 created "2020-11-23" @default.
- W3103183085 creator A5001436196 @default.
- W3103183085 creator A5064802020 @default.
- W3103183085 date "2020-01-01" @default.
- W3103183085 modified "2023-10-18" @default.
- W3103183085 title "Effective Dimension Adaptive Sketching Methods for Faster Regularized Least-Squares Optimization" @default.
- W3103183085 hasPublicationYear "2020" @default.
- W3103183085 type Work @default.
- W3103183085 sameAs 3103183085 @default.
- W3103183085 citedByCount "8" @default.
- W3103183085 countsByYear W31031830852020 @default.
- W3103183085 countsByYear W31031830852021 @default.
- W3103183085 crossrefType "proceedings-article" @default.
- W3103183085 hasAuthorship W3103183085A5001436196 @default.
- W3103183085 hasAuthorship W3103183085A5064802020 @default.
- W3103183085 hasConcept C11413529 @default.
- W3103183085 hasConcept C114614502 @default.
- W3103183085 hasConcept C121332964 @default.
- W3103183085 hasConcept C126255220 @default.
- W3103183085 hasConcept C134306372 @default.
- W3103183085 hasConcept C137836250 @default.
- W3103183085 hasConcept C154945302 @default.
- W3103183085 hasConcept C163716315 @default.
- W3103183085 hasConcept C28826006 @default.
- W3103183085 hasConcept C33676613 @default.
- W3103183085 hasConcept C33923547 @default.
- W3103183085 hasConcept C41008148 @default.
- W3103183085 hasConcept C41608201 @default.
- W3103183085 hasConcept C60292330 @default.
- W3103183085 hasConcept C62520636 @default.
- W3103183085 hasConceptScore W3103183085C11413529 @default.
- W3103183085 hasConceptScore W3103183085C114614502 @default.
- W3103183085 hasConceptScore W3103183085C121332964 @default.
- W3103183085 hasConceptScore W3103183085C126255220 @default.
- W3103183085 hasConceptScore W3103183085C134306372 @default.
- W3103183085 hasConceptScore W3103183085C137836250 @default.
- W3103183085 hasConceptScore W3103183085C154945302 @default.
- W3103183085 hasConceptScore W3103183085C163716315 @default.
- W3103183085 hasConceptScore W3103183085C28826006 @default.
- W3103183085 hasConceptScore W3103183085C33676613 @default.
- W3103183085 hasConceptScore W3103183085C33923547 @default.
- W3103183085 hasConceptScore W3103183085C41008148 @default.
- W3103183085 hasConceptScore W3103183085C41608201 @default.
- W3103183085 hasConceptScore W3103183085C60292330 @default.
- W3103183085 hasConceptScore W3103183085C62520636 @default.
- W3103183085 hasLocation W31031830851 @default.
- W3103183085 hasOpenAccess W3103183085 @default.
- W3103183085 hasPrimaryLocation W31031830851 @default.
- W3103183085 hasRelatedWork W1460549585 @default.
- W3103183085 hasRelatedWork W2117756735 @default.
- W3103183085 hasRelatedWork W2138995291 @default.
- W3103183085 hasRelatedWork W2142642360 @default.
- W3103183085 hasRelatedWork W2185932763 @default.
- W3103183085 hasRelatedWork W2554499514 @default.
- W3103183085 hasRelatedWork W2555830188 @default.
- W3103183085 hasRelatedWork W2883520207 @default.
- W3103183085 hasRelatedWork W2899649398 @default.
- W3103183085 hasRelatedWork W2963060476 @default.
- W3103183085 hasRelatedWork W2970449350 @default.
- W3103183085 hasRelatedWork W2973170454 @default.
- W3103183085 hasRelatedWork W2988331362 @default.
- W3103183085 hasRelatedWork W3034363349 @default.
- W3103183085 hasRelatedWork W3034919836 @default.
- W3103183085 hasRelatedWork W3099937901 @default.
- W3103183085 hasRelatedWork W3133264924 @default.
- W3103183085 hasRelatedWork W3186021868 @default.
- W3103183085 hasRelatedWork W3210624872 @default.
- W3103183085 hasRelatedWork W2570919384 @default.
- W3103183085 hasVolume "33" @default.
- W3103183085 isParatext "false" @default.
- W3103183085 isRetracted "false" @default.
- W3103183085 magId "3103183085" @default.
- W3103183085 workType "article" @default.