FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2892920424> ?p ?o ?g. }

• W2892920424 abstract "Distributed statistical learning problems arise commonly when dealing with large datasets. In this setup, datasets are partitioned over machines, which compute locally, and communicate short messages. Communication is often the bottleneck. In this paper, we study one-step and iterative weighted parameter averaging in statistical linear models under data parallelism. We do linear regression on each machine, send the results to a central server, and take a weighted average of the parameters. Optionally, we iterate, sending back the weighted average and doing local ridge regressions centered at it. How does this work compared to doing linear regression on the full data? Here we study the performance loss in estimation, test error, and confidence interval length in high dimensions, where the number of parameters is comparable to the training data size. We find the performance loss in one-step weighted averaging, and also give results for iterative averaging. We also find that different problems are affected differently by the distributed framework. Estimation error and confidence interval length increase a lot, while prediction error increases much less. We rely on recent results from random matrix theory, where we develop a new calculus of deterministic equivalents as a tool of broader interest." @default.
• W2892920424 created "2018-10-05" @default.
• W2892920424 creator A5031235093 @default.
• W2892920424 creator A5066762360 @default.
• W2892920424 date "2018-09-30" @default.
• W2892920424 modified "2023-09-27" @default.
• W2892920424 title "Distributed linear regression by averaging" @default.
• W2892920424 cites W1586554030 @default.
• W2892920424 cites W1626317705 @default.
• W2892920424 cites W1847890728 @default.
• W2892920424 cites W1968046779 @default.
• W2892920424 cites W199271301 @default.
• W2892920424 cites W2009537245 @default.
• W2892920424 cites W2021883397 @default.
• W2892920424 cites W2034978228 @default.
• W2892920424 cites W2044212084 @default.
• W2892920424 cites W2045391460 @default.
• W2892920424 cites W2060581589 @default.
• W2892920424 cites W2072184935 @default.
• W2892920424 cites W2108475251 @default.
• W2892920424 cites W2109722477 @default.
• W2892920424 cites W2113250264 @default.
• W2892920424 cites W2126119581 @default.
• W2892920424 cites W2133233009 @default.
• W2892920424 cites W2138136390 @default.
• W2892920424 cites W2144839971 @default.
• W2892920424 cites W2154834860 @default.
• W2892920424 cites W2157462866 @default.
• W2892920424 cites W2164278908 @default.
• W2892920424 cites W2165478976 @default.
• W2892920424 cites W2166706236 @default.
• W2892920424 cites W2169713291 @default.
• W2892920424 cites W2173213060 @default.
• W2892920424 cites W2189465200 @default.
• W2892920424 cites W2291895482 @default.
• W2892920424 cites W2553629933 @default.
• W2892920424 cites W2556660792 @default.
• W2892920424 cites W2571425027 @default.
• W2892920424 cites W2595198670 @default.
• W2892920424 cites W2597273636 @default.
• W2892920424 cites W2615069814 @default.
• W2892920424 cites W2734385411 @default.
• W2892920424 cites W2741736887 @default.
• W2892920424 cites W2792513598 @default.
• W2892920424 cites W2801571965 @default.
• W2892920424 cites W2809386967 @default.
• W2892920424 cites W2896498835 @default.
• W2892920424 cites W2896721680 @default.
• W2892920424 cites W2898873098 @default.
• W2892920424 cites W2923749025 @default.
• W2892920424 cites W2923764619 @default.
• W2892920424 cites W2949198759 @default.
• W2892920424 cites W2951572058 @default.
• W2892920424 cites W2963020641 @default.
• W2892920424 cites W2963051744 @default.
• W2892920424 cites W2963094815 @default.
• W2892920424 cites W2963106396 @default.
• W2892920424 cites W2963126228 @default.
• W2892920424 cites W2963422939 @default.
• W2892920424 cites W2963788280 @default.
• W2892920424 cites W2963992805 @default.
• W2892920424 cites W2964166170 @default.
• W2892920424 cites W2964274100 @default.
• W2892920424 cites W3007034419 @default.
• W2892920424 cites W3101665129 @default.
• W2892920424 cites W3103096700 @default.
• W2892920424 cites W3103692684 @default.
• W2892920424 cites W3103741939 @default.
• W2892920424 cites W3133118450 @default.
• W2892920424 cites W615589970 @default.
• W2892920424 doi "https://doi.org/10.48550/arxiv.1810.00412" @default.
• W2892920424 hasPublicationYear "2018" @default.
• W2892920424 type Work @default.
• W2892920424 sameAs 2892920424 @default.
• W2892920424 citedByCount "11" @default.
• W2892920424 countsByYear W28929204242019 @default.
• W2892920424 countsByYear W28929204242020 @default.
• W2892920424 countsByYear W28929204242021 @default.
• W2892920424 crossrefType "posted-content" @default.
• W2892920424 hasAuthorship W2892920424A5031235093 @default.
• W2892920424 hasAuthorship W2892920424A5066762360 @default.
• W2892920424 hasBestOaLocation W28929204241 @default.
• W2892920424 hasConcept C105795698 @default.
• W2892920424 hasConcept C11413529 @default.
• W2892920424 hasConcept C149635348 @default.
• W2892920424 hasConcept C152877465 @default.
• W2892920424 hasConcept C163175372 @default.
• W2892920424 hasConcept C203233044 @default.
• W2892920424 hasConcept C2780513914 @default.
• W2892920424 hasConcept C33923547 @default.
• W2892920424 hasConcept C41008148 @default.
• W2892920424 hasConcept C44249647 @default.
• W2892920424 hasConcept C48921125 @default.
• W2892920424 hasConcept C83546350 @default.
• W2892920424 hasConceptScore W2892920424C105795698 @default.
• W2892920424 hasConceptScore W2892920424C11413529 @default.
• W2892920424 hasConceptScore W2892920424C149635348 @default.
• W2892920424 hasConceptScore W2892920424C152877465 @default.
• W2892920424 hasConceptScore W2892920424C163175372 @default.
• W2892920424 hasConceptScore W2892920424C203233044 @default.

• next