Matches in SemOpenAlex for { <https://semopenalex.org/work/W1796770224> ?p ?o ?g. }
- W1796770224 abstract "Partition functions arise in a variety of settings, including conditional random fields, logistic regression, and latent gaussian models. In this paper, we consider semistochastic quadratic bound (SQB) methods for maximum likelihood inference based on partition function optimization. Batch methods based on the quadratic bound were recently proposed for this class of problems, and performed favorably in comparison to state-of-the-art techniques. Semistochastic methods fall in between batch algorithms, which use all the data, and stochastic gradient type methods, which use small random selections at each iteration. We build semistochastic quadratic bound-based methods, and prove both global convergence (to a stationary point) under very weak assumptions, and linear convergence rate under stronger assumptions on the objective. To make the proposed methods faster and more stable, we consider inexact subproblem minimization and batch-size selection schemes. The efficacy of SQB methods is demonstrated via comparison with several state-of-the-art techniques on commonly used datasets." @default.
- W1796770224 created "2016-06-24" @default.
- W1796770224 creator A5006452373 @default.
- W1796770224 creator A5046671967 @default.
- W1796770224 creator A5068077704 @default.
- W1796770224 creator A5091863515 @default.
- W1796770224 date "2013-09-05" @default.
- W1796770224 modified "2023-10-01" @default.
- W1796770224 title "Semistochastic Quadratic Bound Methods" @default.
- W1796770224 cites W1491622225 @default.
- W1796770224 cites W1568229137 @default.
- W1796770224 cites W1890144115 @default.
- W1796770224 cites W196761320 @default.
- W1796770224 cites W1971942712 @default.
- W1796770224 cites W1983916623 @default.
- W1796770224 cites W1991083751 @default.
- W1796770224 cites W1992208280 @default.
- W1796770224 cites W1994616650 @default.
- W1796770224 cites W2040870580 @default.
- W1796770224 cites W2061570747 @default.
- W1796770224 cites W2061863621 @default.
- W1796770224 cites W2064133327 @default.
- W1796770224 cites W2086161653 @default.
- W1796770224 cites W2096840748 @default.
- W1796770224 cites W2101836039 @default.
- W1796770224 cites W2102017903 @default.
- W1796770224 cites W2102486516 @default.
- W1796770224 cites W2105875671 @default.
- W1796770224 cites W2107438106 @default.
- W1796770224 cites W2111200589 @default.
- W1796770224 cites W2114424556 @default.
- W1796770224 cites W2119374337 @default.
- W1796770224 cites W2129113961 @default.
- W1796770224 cites W2129142203 @default.
- W1796770224 cites W2130984546 @default.
- W1796770224 cites W2137515395 @default.
- W1796770224 cites W2145832734 @default.
- W1796770224 cites W2169713291 @default.
- W1796770224 cites W2316564661 @default.
- W1796770224 cites W2762578791 @default.
- W1796770224 cites W2951833684 @default.
- W1796770224 cites W2952594493 @default.
- W1796770224 cites W71081281 @default.
- W1796770224 hasPublicationYear "2013" @default.
- W1796770224 type Work @default.
- W1796770224 sameAs 1796770224 @default.
- W1796770224 citedByCount "0" @default.
- W1796770224 crossrefType "posted-content" @default.
- W1796770224 hasAuthorship W1796770224A5006452373 @default.
- W1796770224 hasAuthorship W1796770224A5046671967 @default.
- W1796770224 hasAuthorship W1796770224A5068077704 @default.
- W1796770224 hasAuthorship W1796770224A5091863515 @default.
- W1796770224 hasConcept C114614502 @default.
- W1796770224 hasConcept C121332964 @default.
- W1796770224 hasConcept C126255220 @default.
- W1796770224 hasConcept C129844170 @default.
- W1796770224 hasConcept C134306372 @default.
- W1796770224 hasConcept C154945302 @default.
- W1796770224 hasConcept C163716315 @default.
- W1796770224 hasConcept C189237950 @default.
- W1796770224 hasConcept C2524010 @default.
- W1796770224 hasConcept C2776214188 @default.
- W1796770224 hasConcept C28826006 @default.
- W1796770224 hasConcept C33923547 @default.
- W1796770224 hasConcept C41008148 @default.
- W1796770224 hasConcept C42812 @default.
- W1796770224 hasConcept C62520636 @default.
- W1796770224 hasConcept C77553402 @default.
- W1796770224 hasConceptScore W1796770224C114614502 @default.
- W1796770224 hasConceptScore W1796770224C121332964 @default.
- W1796770224 hasConceptScore W1796770224C126255220 @default.
- W1796770224 hasConceptScore W1796770224C129844170 @default.
- W1796770224 hasConceptScore W1796770224C134306372 @default.
- W1796770224 hasConceptScore W1796770224C154945302 @default.
- W1796770224 hasConceptScore W1796770224C163716315 @default.
- W1796770224 hasConceptScore W1796770224C189237950 @default.
- W1796770224 hasConceptScore W1796770224C2524010 @default.
- W1796770224 hasConceptScore W1796770224C2776214188 @default.
- W1796770224 hasConceptScore W1796770224C28826006 @default.
- W1796770224 hasConceptScore W1796770224C33923547 @default.
- W1796770224 hasConceptScore W1796770224C41008148 @default.
- W1796770224 hasConceptScore W1796770224C42812 @default.
- W1796770224 hasConceptScore W1796770224C62520636 @default.
- W1796770224 hasConceptScore W1796770224C77553402 @default.
- W1796770224 hasLocation W17967702241 @default.
- W1796770224 hasOpenAccess W1796770224 @default.
- W1796770224 hasPrimaryLocation W17967702241 @default.
- W1796770224 hasRelatedWork W1972950354 @default.
- W1796770224 hasRelatedWork W2060522901 @default.
- W1796770224 hasRelatedWork W2095999217 @default.
- W1796770224 hasRelatedWork W2145832734 @default.
- W1796770224 hasRelatedWork W2168601171 @default.
- W1796770224 hasRelatedWork W2233018254 @default.
- W1796770224 hasRelatedWork W2331993650 @default.
- W1796770224 hasRelatedWork W2555906060 @default.
- W1796770224 hasRelatedWork W2770391550 @default.
- W1796770224 hasRelatedWork W2890661317 @default.
- W1796770224 hasRelatedWork W2921349977 @default.
- W1796770224 hasRelatedWork W2960723499 @default.
- W1796770224 hasRelatedWork W2964140893 @default.