Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384613897> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4384613897 abstract "This paper considers the problem of understanding the behavior of a general class of accelerated gradient methods on smooth nonconvex functions. Motivated by some recent works that have proposed effective algorithms, based on Polyak's heavy ball method and the Nesterov accelerated gradient method, to achieve convergence to a local minimum of nonconvex functions, this work proposes a broad class of Nesterov-type accelerated methods and puts forth a rigorous study of these methods encompassing the escape from saddle-points and convergence to local minima through a both asymptotic and a non-asymptotic analysis. In the asymptotic regime, this paper answers an open question of whether Nesterov's accelerated gradient method (NAG) with variable momentum parameter avoids strict saddle points almost surely. This work also develops two metrics of asymptotic rate of convergence and divergence, and evaluates these two metrics for several popular standard accelerated methods such as the NAG, and Nesterov's accelerated gradient with constant momentum (NCM) near strict saddle points. In the local regime, this work provides an analysis that leads to the linear exit time estimates from strict saddle neighborhoods for trajectories of these accelerated methods as well the necessary conditions for the existence of such trajectories. Finally, this work studies a sub-class of accelerated methods that can converge in convex neighborhoods of nonconvex functions with a near optimal rate to a local minima and at the same time this sub-class offers superior saddle-escape behavior compared to that of NAG." @default.
- W4384613897 created "2023-07-18" @default.
- W4384613897 creator A5028718006 @default.
- W4384613897 creator A5034370705 @default.
- W4384613897 creator A5061333650 @default.
- W4384613897 date "2023-07-13" @default.
- W4384613897 modified "2023-09-23" @default.
- W4384613897 title "Accelerated gradient methods for nonconvex optimization: Escape trajectories from strict saddle points and convergence to local minima" @default.
- W4384613897 doi "https://doi.org/10.48550/arxiv.2307.07030" @default.
- W4384613897 hasPublicationYear "2023" @default.
- W4384613897 type Work @default.
- W4384613897 citedByCount "0" @default.
- W4384613897 crossrefType "posted-content" @default.
- W4384613897 hasAuthorship W4384613897A5028718006 @default.
- W4384613897 hasAuthorship W4384613897A5034370705 @default.
- W4384613897 hasAuthorship W4384613897A5061333650 @default.
- W4384613897 hasBestOaLocation W43846138971 @default.
- W4384613897 hasConcept C112680207 @default.
- W4384613897 hasConcept C122041747 @default.
- W4384613897 hasConcept C126255220 @default.
- W4384613897 hasConcept C127162648 @default.
- W4384613897 hasConcept C134306372 @default.
- W4384613897 hasConcept C145446738 @default.
- W4384613897 hasConcept C162324750 @default.
- W4384613897 hasConcept C186633575 @default.
- W4384613897 hasConcept C2524010 @default.
- W4384613897 hasConcept C2681867 @default.
- W4384613897 hasConcept C2777127463 @default.
- W4384613897 hasConcept C2777303404 @default.
- W4384613897 hasConcept C28826006 @default.
- W4384613897 hasConcept C31258907 @default.
- W4384613897 hasConcept C33923547 @default.
- W4384613897 hasConcept C41008148 @default.
- W4384613897 hasConcept C50522688 @default.
- W4384613897 hasConcept C57869625 @default.
- W4384613897 hasConceptScore W4384613897C112680207 @default.
- W4384613897 hasConceptScore W4384613897C122041747 @default.
- W4384613897 hasConceptScore W4384613897C126255220 @default.
- W4384613897 hasConceptScore W4384613897C127162648 @default.
- W4384613897 hasConceptScore W4384613897C134306372 @default.
- W4384613897 hasConceptScore W4384613897C145446738 @default.
- W4384613897 hasConceptScore W4384613897C162324750 @default.
- W4384613897 hasConceptScore W4384613897C186633575 @default.
- W4384613897 hasConceptScore W4384613897C2524010 @default.
- W4384613897 hasConceptScore W4384613897C2681867 @default.
- W4384613897 hasConceptScore W4384613897C2777127463 @default.
- W4384613897 hasConceptScore W4384613897C2777303404 @default.
- W4384613897 hasConceptScore W4384613897C28826006 @default.
- W4384613897 hasConceptScore W4384613897C31258907 @default.
- W4384613897 hasConceptScore W4384613897C33923547 @default.
- W4384613897 hasConceptScore W4384613897C41008148 @default.
- W4384613897 hasConceptScore W4384613897C50522688 @default.
- W4384613897 hasConceptScore W4384613897C57869625 @default.
- W4384613897 hasLocation W43846138971 @default.
- W4384613897 hasOpenAccess W4384613897 @default.
- W4384613897 hasPrimaryLocation W43846138971 @default.
- W4384613897 hasRelatedWork W2051848413 @default.
- W4384613897 hasRelatedWork W2352701666 @default.
- W4384613897 hasRelatedWork W2592651140 @default.
- W4384613897 hasRelatedWork W2771077211 @default.
- W4384613897 hasRelatedWork W2952233565 @default.
- W4384613897 hasRelatedWork W2963092340 @default.
- W4384613897 hasRelatedWork W2963586744 @default.
- W4384613897 hasRelatedWork W3003619226 @default.
- W4384613897 hasRelatedWork W3046748628 @default.
- W4384613897 hasRelatedWork W32532385 @default.
- W4384613897 isParatext "false" @default.
- W4384613897 isRetracted "false" @default.
- W4384613897 workType "article" @default.