Matches in SemOpenAlex for { <https://semopenalex.org/work/W2894101493> ?p ?o ?g. }
- W2894101493 abstract "Several first order stochastic optimization methods commonly used in the Euclidean domain such as stochastic gradient descent (SGD), accelerated gradient descent or variance reduced methods have already been adapted to certain Riemannian settings. However, some of the most popular of these optimization tools - namely Adam , Adagrad and the more recent Amsgrad - remain to be generalized to Riemannian manifolds. We discuss the difficulty of generalizing such adaptive schemes to the most agnostic Riemannian setting, and then provide algorithms and convergence proofs for geodesically convex objectives in the particular case of a product of Riemannian manifolds, in which adaptivity is implemented across manifolds in the cartesian product. Our generalization is tight in the sense that choosing the Euclidean space as Riemannian manifold yields the same algorithms and regret bounds as those that were already known for the standard algorithms. Experimentally, we show faster convergence and to a lower train loss value for Riemannian adaptive methods over their corresponding baselines on the realistic task of embedding the WordNet taxonomy in the Poincare ball." @default.
- W2894101493 created "2018-10-05" @default.
- W2894101493 creator A5037156810 @default.
- W2894101493 creator A5064290471 @default.
- W2894101493 date "2018-10-01" @default.
- W2894101493 modified "2023-09-23" @default.
- W2894101493 title "Riemannian Adaptive Optimization Methods" @default.
- W2894101493 cites W148638337 @default.
- W2894101493 cites W1587069887 @default.
- W2894101493 cites W2001070843 @default.
- W2894101493 cites W2037253724 @default.
- W2894101493 cites W2055639053 @default.
- W2894101493 cites W2074763854 @default.
- W2894101493 cites W2102381086 @default.
- W2894101493 cites W2110529144 @default.
- W2894101493 cites W2146502635 @default.
- W2894101493 cites W2250539671 @default.
- W2894101493 cites W2307703883 @default.
- W2894101493 cites W2549348365 @default.
- W2894101493 cites W2753579054 @default.
- W2894101493 cites W2774598382 @default.
- W2894101493 cites W2785523195 @default.
- W2894101493 cites W2798738034 @default.
- W2894101493 cites W2807088554 @default.
- W2894101493 cites W2890645261 @default.
- W2894101493 cites W2962774730 @default.
- W2894101493 cites W2962836826 @default.
- W2894101493 cites W2962905348 @default.
- W2894101493 cites W2962918699 @default.
- W2894101493 cites W2962936818 @default.
- W2894101493 cites W2963849380 @default.
- W2894101493 cites W2964109691 @default.
- W2894101493 cites W2964121744 @default.
- W2894101493 cites W2964297142 @default.
- W2894101493 cites W3102317997 @default.
- W2894101493 cites W6908809 @default.
- W2894101493 hasPublicationYear "2018" @default.
- W2894101493 type Work @default.
- W2894101493 sameAs 2894101493 @default.
- W2894101493 citedByCount "6" @default.
- W2894101493 countsByYear W28941014932019 @default.
- W2894101493 countsByYear W28941014932020 @default.
- W2894101493 countsByYear W28941014932021 @default.
- W2894101493 crossrefType "posted-content" @default.
- W2894101493 hasAuthorship W2894101493A5037156810 @default.
- W2894101493 hasAuthorship W2894101493A5064290471 @default.
- W2894101493 hasConcept C108710211 @default.
- W2894101493 hasConcept C12520029 @default.
- W2894101493 hasConcept C126255220 @default.
- W2894101493 hasConcept C153258448 @default.
- W2894101493 hasConcept C154945302 @default.
- W2894101493 hasConcept C181104567 @default.
- W2894101493 hasConcept C186450821 @default.
- W2894101493 hasConcept C195065555 @default.
- W2894101493 hasConcept C202444582 @default.
- W2894101493 hasConcept C206688291 @default.
- W2894101493 hasConcept C2524010 @default.
- W2894101493 hasConcept C2779593128 @default.
- W2894101493 hasConcept C28826006 @default.
- W2894101493 hasConcept C33923547 @default.
- W2894101493 hasConcept C41008148 @default.
- W2894101493 hasConcept C4306028 @default.
- W2894101493 hasConcept C50644808 @default.
- W2894101493 hasConceptScore W2894101493C108710211 @default.
- W2894101493 hasConceptScore W2894101493C12520029 @default.
- W2894101493 hasConceptScore W2894101493C126255220 @default.
- W2894101493 hasConceptScore W2894101493C153258448 @default.
- W2894101493 hasConceptScore W2894101493C154945302 @default.
- W2894101493 hasConceptScore W2894101493C181104567 @default.
- W2894101493 hasConceptScore W2894101493C186450821 @default.
- W2894101493 hasConceptScore W2894101493C195065555 @default.
- W2894101493 hasConceptScore W2894101493C202444582 @default.
- W2894101493 hasConceptScore W2894101493C206688291 @default.
- W2894101493 hasConceptScore W2894101493C2524010 @default.
- W2894101493 hasConceptScore W2894101493C2779593128 @default.
- W2894101493 hasConceptScore W2894101493C28826006 @default.
- W2894101493 hasConceptScore W2894101493C33923547 @default.
- W2894101493 hasConceptScore W2894101493C41008148 @default.
- W2894101493 hasConceptScore W2894101493C4306028 @default.
- W2894101493 hasConceptScore W2894101493C50644808 @default.
- W2894101493 hasOpenAccess W2894101493 @default.
- W2894101493 hasRelatedWork W1632150706 @default.
- W2894101493 hasRelatedWork W165780948 @default.
- W2894101493 hasRelatedWork W2146971896 @default.
- W2894101493 hasRelatedWork W2194775991 @default.
- W2894101493 hasRelatedWork W2199357291 @default.
- W2894101493 hasRelatedWork W2766367585 @default.
- W2894101493 hasRelatedWork W2807213139 @default.
- W2894101493 hasRelatedWork W2893613551 @default.
- W2894101493 hasRelatedWork W2962918699 @default.
- W2894101493 hasRelatedWork W2962936818 @default.
- W2894101493 hasRelatedWork W2963849380 @default.
- W2894101493 hasRelatedWork W2964050265 @default.
- W2894101493 hasRelatedWork W3005669520 @default.
- W2894101493 hasRelatedWork W3102317997 @default.
- W2894101493 hasRelatedWork W3103171021 @default.
- W2894101493 hasRelatedWork W3123332075 @default.
- W2894101493 hasRelatedWork W3158914048 @default.
- W2894101493 hasRelatedWork W3172480780 @default.
- W2894101493 hasRelatedWork W3195473549 @default.