FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2548231749> ?p ?o ?g. }

• W2548231749 abstract "Recently, there has been growing interest in developing optimization methods for solving large-scale machine learning problems. Most of these problems boil down to the problem of minimizing an average of a finite set of smooth and strongly convex functions where the number of functions $n$ is large. Gradient descent method (GD) is successful in minimizing convex problems at a fast linear rate; however, it is not applicable to the considered large-scale optimization setting because of the high computational complexity. Incremental methods resolve this drawback of gradient methods by replacing the required gradient for the descent direction with an incremental gradient approximation. They operate by evaluating one gradient per iteration and executing the average of the $n$ available gradients as a gradient approximate. Although, incremental methods reduce the computational cost of GD, their convergence rates do not justify their advantage relative to GD in terms of the total number of gradient evaluations until convergence. In this paper, we introduce a Double Incremental Aggregated Gradient method (DIAG) that computes the gradient of only one function at each iteration, which is chosen based on a cyclic scheme, and uses the aggregated average gradient of all the functions to approximate the full gradient. The iterates of the proposed DIAG method uses averages of both iterates and gradients in oppose to classic incremental methods that utilize gradient averages but do not utilize iterate averages. We prove that not only the proposed DIAG method converges linearly to the optimal solution, but also its linear convergence factor justifies the advantage of incremental methods on GD. In particular, we prove that the worst case performance of DIAG is better than the worst case performance of GD." @default.
• W2548231749 created "2016-11-11" @default.
• W2548231749 creator A5008417632 @default.
• W2548231749 creator A5034370705 @default.
• W2548231749 creator A5078862959 @default.
• W2548231749 date "2016-11-01" @default.
• W2548231749 modified "2023-09-22" @default.
• W2548231749 title "Surpassing Gradient Descent Provably: A Cyclic Incremental Method with Linear Convergence Rate" @default.
• W2548231749 cites W114517082 @default.
• W2548231749 cites W1488435683 @default.
• W2548231749 cites W1523661875 @default.
• W2548231749 cites W1788292158 @default.
• W2548231749 cites W1939652453 @default.
• W2548231749 cites W1970997001 @default.
• W2548231749 cites W1973754217 @default.
• W2548231749 cites W1988795359 @default.
• W2548231749 cites W1994616650 @default.
• W2548231749 cites W2012812921 @default.
• W2548231749 cites W2017938640 @default.
• W2548231749 cites W2020909452 @default.
• W2548231749 cites W2030811966 @default.
• W2548231749 cites W2047152541 @default.
• W2548231749 cites W2049659086 @default.
• W2548231749 cites W2061863621 @default.
• W2548231749 cites W2073750241 @default.
• W2548231749 cites W2098741260 @default.
• W2548231749 cites W2107438106 @default.
• W2548231749 cites W2116222104 @default.
• W2548231749 cites W2118545728 @default.
• W2548231749 cites W2120717492 @default.
• W2548231749 cites W2124541940 @default.
• W2548231749 cites W2141788746 @default.
• W2548231749 cites W2216483973 @default.
• W2548231749 cites W2499557797 @default.
• W2548231749 cites W2640723840 @default.
• W2548231749 cites W2951120012 @default.
• W2548231749 cites W2952215077 @default.
• W2548231749 cites W3104398353 @default.
• W2548231749 cites W3123607434 @default.
• W2548231749 hasPublicationYear "2016" @default.
• W2548231749 type Work @default.
• W2548231749 sameAs 2548231749 @default.
• W2548231749 citedByCount "3" @default.
• W2548231749 countsByYear W25482317492017 @default.
• W2548231749 countsByYear W25482317492018 @default.
• W2548231749 crossrefType "posted-content" @default.
• W2548231749 hasAuthorship W2548231749A5008417632 @default.
• W2548231749 hasAuthorship W2548231749A5034370705 @default.
• W2548231749 hasAuthorship W2548231749A5078862959 @default.
• W2548231749 hasConcept C10494615 @default.
• W2548231749 hasConcept C112680207 @default.
• W2548231749 hasConcept C11413529 @default.
• W2548231749 hasConcept C115680565 @default.
• W2548231749 hasConcept C119857082 @default.
• W2548231749 hasConcept C126255220 @default.
• W2548231749 hasConcept C127162648 @default.
• W2548231749 hasConcept C134306372 @default.
• W2548231749 hasConcept C14036430 @default.
• W2548231749 hasConcept C140479938 @default.
• W2548231749 hasConcept C145446738 @default.
• W2548231749 hasConcept C153258448 @default.
• W2548231749 hasConcept C162324750 @default.
• W2548231749 hasConcept C2524010 @default.
• W2548231749 hasConcept C2777303404 @default.
• W2548231749 hasConcept C28826006 @default.
• W2548231749 hasConcept C31258907 @default.
• W2548231749 hasConcept C33923547 @default.
• W2548231749 hasConcept C41008148 @default.
• W2548231749 hasConcept C50522688 @default.
• W2548231749 hasConcept C50644808 @default.
• W2548231749 hasConcept C57869625 @default.
• W2548231749 hasConcept C78458016 @default.
• W2548231749 hasConcept C86803240 @default.
• W2548231749 hasConceptScore W2548231749C10494615 @default.
• W2548231749 hasConceptScore W2548231749C112680207 @default.
• W2548231749 hasConceptScore W2548231749C11413529 @default.
• W2548231749 hasConceptScore W2548231749C115680565 @default.
• W2548231749 hasConceptScore W2548231749C119857082 @default.
• W2548231749 hasConceptScore W2548231749C126255220 @default.
• W2548231749 hasConceptScore W2548231749C127162648 @default.
• W2548231749 hasConceptScore W2548231749C134306372 @default.
• W2548231749 hasConceptScore W2548231749C14036430 @default.
• W2548231749 hasConceptScore W2548231749C140479938 @default.
• W2548231749 hasConceptScore W2548231749C145446738 @default.
• W2548231749 hasConceptScore W2548231749C153258448 @default.
• W2548231749 hasConceptScore W2548231749C162324750 @default.
• W2548231749 hasConceptScore W2548231749C2524010 @default.
• W2548231749 hasConceptScore W2548231749C2777303404 @default.
• W2548231749 hasConceptScore W2548231749C28826006 @default.
• W2548231749 hasConceptScore W2548231749C31258907 @default.
• W2548231749 hasConceptScore W2548231749C33923547 @default.
• W2548231749 hasConceptScore W2548231749C41008148 @default.
• W2548231749 hasConceptScore W2548231749C50522688 @default.
• W2548231749 hasConceptScore W2548231749C50644808 @default.
• W2548231749 hasConceptScore W2548231749C57869625 @default.
• W2548231749 hasConceptScore W2548231749C78458016 @default.
• W2548231749 hasConceptScore W2548231749C86803240 @default.
• W2548231749 hasOpenAccess W2548231749 @default.
• W2548231749 hasRelatedWork W2098741260 @default.
• W2548231749 hasRelatedWork W2499557797 @default.

• next