Matches in SemOpenAlex for { <https://semopenalex.org/work/W4299446532> ?p ?o ?g. }
Showing items 1 to 72 of
72
with 100 items per page.
- W4299446532 abstract "We study the trade-offs between convergence rate and robustness to gradient errors in designing a first-order algorithm. We focus on gradient descent (GD) and accelerated gradient (AG) methods for minimizing strongly convex functions when the gradient has random errors in the form of additive white noise. With gradient errors, the function values of the iterates need not converge to the optimal value; hence, we define the robustness of an algorithm to noise as the asymptotic expected suboptimality of the iterate sequence to input noise power. For this robustness measure, we provide exact expressions for the quadratic case using tools from robust control theory and tight upper bounds for the smooth strongly convex case using Lyapunov functions certified through matrix inequalities. We use these characterizations within an optimization problem which selects parameters of each algorithm to achieve a particular trade-off between rate and robustness. Our results show that AG can achieve acceleration while being more robust to random gradient errors. This behavior is quite different than previously reported in the deterministic gradient noise setting. We also establish some connections between the robustness of an algorithm and how quickly it can converge back to the optimal solution if it is perturbed from the optimal point with deterministic noise. Our framework also leads to practical algorithms that can perform better than other state-of-the-art methods in the presence of random gradient noise." @default.
- W4299446532 created "2022-10-02" @default.
- W4299446532 creator A5016741616 @default.
- W4299446532 creator A5034370705 @default.
- W4299446532 creator A5067307504 @default.
- W4299446532 creator A5076553434 @default.
- W4299446532 date "2018-05-27" @default.
- W4299446532 modified "2023-10-16" @default.
- W4299446532 title "Robust Accelerated Gradient Methods for Smooth Strongly Convex Functions" @default.
- W4299446532 doi "https://doi.org/10.48550/arxiv.1805.10579" @default.
- W4299446532 hasPublicationYear "2018" @default.
- W4299446532 type Work @default.
- W4299446532 citedByCount "0" @default.
- W4299446532 crossrefType "posted-content" @default.
- W4299446532 hasAuthorship W4299446532A5016741616 @default.
- W4299446532 hasAuthorship W4299446532A5034370705 @default.
- W4299446532 hasAuthorship W4299446532A5067307504 @default.
- W4299446532 hasAuthorship W4299446532A5076553434 @default.
- W4299446532 hasBestOaLocation W42994465321 @default.
- W4299446532 hasConcept C104317684 @default.
- W4299446532 hasConcept C112680207 @default.
- W4299446532 hasConcept C119857082 @default.
- W4299446532 hasConcept C126255220 @default.
- W4299446532 hasConcept C127162648 @default.
- W4299446532 hasConcept C134306372 @default.
- W4299446532 hasConcept C140479938 @default.
- W4299446532 hasConcept C145446738 @default.
- W4299446532 hasConcept C153258448 @default.
- W4299446532 hasConcept C185592680 @default.
- W4299446532 hasConcept C2524010 @default.
- W4299446532 hasConcept C31258907 @default.
- W4299446532 hasConcept C33923547 @default.
- W4299446532 hasConcept C41008148 @default.
- W4299446532 hasConcept C50644808 @default.
- W4299446532 hasConcept C55493867 @default.
- W4299446532 hasConcept C57869625 @default.
- W4299446532 hasConcept C63479239 @default.
- W4299446532 hasConceptScore W4299446532C104317684 @default.
- W4299446532 hasConceptScore W4299446532C112680207 @default.
- W4299446532 hasConceptScore W4299446532C119857082 @default.
- W4299446532 hasConceptScore W4299446532C126255220 @default.
- W4299446532 hasConceptScore W4299446532C127162648 @default.
- W4299446532 hasConceptScore W4299446532C134306372 @default.
- W4299446532 hasConceptScore W4299446532C140479938 @default.
- W4299446532 hasConceptScore W4299446532C145446738 @default.
- W4299446532 hasConceptScore W4299446532C153258448 @default.
- W4299446532 hasConceptScore W4299446532C185592680 @default.
- W4299446532 hasConceptScore W4299446532C2524010 @default.
- W4299446532 hasConceptScore W4299446532C31258907 @default.
- W4299446532 hasConceptScore W4299446532C33923547 @default.
- W4299446532 hasConceptScore W4299446532C41008148 @default.
- W4299446532 hasConceptScore W4299446532C50644808 @default.
- W4299446532 hasConceptScore W4299446532C55493867 @default.
- W4299446532 hasConceptScore W4299446532C57869625 @default.
- W4299446532 hasConceptScore W4299446532C63479239 @default.
- W4299446532 hasLocation W42994465321 @default.
- W4299446532 hasLocation W42994465322 @default.
- W4299446532 hasOpenAccess W4299446532 @default.
- W4299446532 hasPrimaryLocation W42994465321 @default.
- W4299446532 hasRelatedWork W2137731592 @default.
- W4299446532 hasRelatedWork W2548231749 @default.
- W4299446532 hasRelatedWork W2570633406 @default.
- W4299446532 hasRelatedWork W2788611447 @default.
- W4299446532 hasRelatedWork W2893725511 @default.
- W4299446532 hasRelatedWork W2952586095 @default.
- W4299446532 hasRelatedWork W2980820424 @default.
- W4299446532 hasRelatedWork W4289465566 @default.
- W4299446532 hasRelatedWork W4298016328 @default.
- W4299446532 hasRelatedWork W4300547822 @default.
- W4299446532 isParatext "false" @default.
- W4299446532 isRetracted "false" @default.
- W4299446532 workType "article" @default.