FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2950525610> ?p ?o ?g. }

• W2950525610 abstract "Mini-batch gradient descent based methods are the de facto algorithms for training neural network architectures today. We introduce a mini-batch selection strategy based on submodular function maximization. Our novel submodular formulation captures the informativeness of each sample and diversity of the whole subset. We design an efficient, greedy algorithm which can give high-quality solutions to this NP-hard combinatorial optimization problem. Our extensive experiments on standard datasets show that the deep models trained using the proposed batch selection strategy provide better generalization than Stochastic Gradient Descent as well as a popular baseline sampling strategy across different learning rates, batch sizes, and distance metrics." @default.
• W2950525610 created "2019-06-27" @default.
• W2950525610 creator A5024380274 @default.
• W2950525610 creator A5035656808 @default.
• W2950525610 creator A5038020125 @default.
• W2950525610 creator A5060320135 @default.
• W2950525610 date "2019-06-20" @default.
• W2950525610 modified "2023-10-17" @default.
• W2950525610 title "Submodular Batch Selection for Training Deep Neural Networks" @default.
• W2950525610 cites W1522301498 @default.
• W2950525610 cites W1680189815 @default.
• W2950525610 cites W1757790940 @default.
• W2950525610 cites W1842094663 @default.
• W2950525610 cites W1898824936 @default.
• W2950525610 cites W1912128066 @default.
• W2950525610 cites W2072566913 @default.
• W2950525610 cites W2100729440 @default.
• W2950525610 cites W2107438106 @default.
• W2950525610 cites W2113097878 @default.
• W2950525610 cites W2144933361 @default.
• W2950525610 cites W2159416808 @default.
• W2950525610 cites W2170735271 @default.
• W2950525610 cites W2174940656 @default.
• W2950525610 cites W2177410802 @default.
• W2950525610 cites W2194775991 @default.
• W2950525610 cites W2335728318 @default.
• W2950525610 cites W2621265919 @default.
• W2950525610 cites W2794302998 @default.
• W2950525610 cites W2795730307 @default.
• W2950525610 cites W2883919314 @default.
• W2950525610 cites W2884265051 @default.
• W2950525610 cites W2899771611 @default.
• W2950525610 cites W2917620156 @default.
• W2950525610 cites W2949117887 @default.
• W2950525610 cites W2952575824 @default.
• W2950525610 cites W2962685744 @default.
• W2950525610 cites W2963476860 @default.
• W2950525610 cites W3118608800 @default.
• W2950525610 doi "https://doi.org/10.48550/arxiv.1906.08771" @default.
• W2950525610 hasPublicationYear "2019" @default.
• W2950525610 type Work @default.
• W2950525610 sameAs 2950525610 @default.
• W2950525610 citedByCount "0" @default.
• W2950525610 crossrefType "posted-content" @default.
• W2950525610 hasAuthorship W2950525610A5024380274 @default.
• W2950525610 hasAuthorship W2950525610A5035656808 @default.
• W2950525610 hasAuthorship W2950525610A5038020125 @default.
• W2950525610 hasAuthorship W2950525610A5060320135 @default.
• W2950525610 hasBestOaLocation W29505256101 @default.
• W2950525610 hasConcept C111368507 @default.
• W2950525610 hasConcept C11413529 @default.
• W2950525610 hasConcept C119857082 @default.
• W2950525610 hasConcept C126255220 @default.
• W2950525610 hasConcept C12725497 @default.
• W2950525610 hasConcept C127313418 @default.
• W2950525610 hasConcept C134306372 @default.
• W2950525610 hasConcept C14036430 @default.
• W2950525610 hasConcept C153258448 @default.
• W2950525610 hasConcept C154945302 @default.
• W2950525610 hasConcept C172658912 @default.
• W2950525610 hasConcept C177148314 @default.
• W2950525610 hasConcept C178621042 @default.
• W2950525610 hasConcept C185592680 @default.
• W2950525610 hasConcept C198531522 @default.
• W2950525610 hasConcept C199360897 @default.
• W2950525610 hasConcept C206688291 @default.
• W2950525610 hasConcept C2776330181 @default.
• W2950525610 hasConcept C33923547 @default.
• W2950525610 hasConcept C41008148 @default.
• W2950525610 hasConcept C43617362 @default.
• W2950525610 hasConcept C50644808 @default.
• W2950525610 hasConcept C51823790 @default.
• W2950525610 hasConcept C78458016 @default.
• W2950525610 hasConcept C81917197 @default.
• W2950525610 hasConcept C86803240 @default.
• W2950525610 hasConceptScore W2950525610C111368507 @default.
• W2950525610 hasConceptScore W2950525610C11413529 @default.
• W2950525610 hasConceptScore W2950525610C119857082 @default.
• W2950525610 hasConceptScore W2950525610C126255220 @default.
• W2950525610 hasConceptScore W2950525610C12725497 @default.
• W2950525610 hasConceptScore W2950525610C127313418 @default.
• W2950525610 hasConceptScore W2950525610C134306372 @default.
• W2950525610 hasConceptScore W2950525610C14036430 @default.
• W2950525610 hasConceptScore W2950525610C153258448 @default.
• W2950525610 hasConceptScore W2950525610C154945302 @default.
• W2950525610 hasConceptScore W2950525610C172658912 @default.
• W2950525610 hasConceptScore W2950525610C177148314 @default.
• W2950525610 hasConceptScore W2950525610C178621042 @default.
• W2950525610 hasConceptScore W2950525610C185592680 @default.
• W2950525610 hasConceptScore W2950525610C198531522 @default.
• W2950525610 hasConceptScore W2950525610C199360897 @default.
• W2950525610 hasConceptScore W2950525610C206688291 @default.
• W2950525610 hasConceptScore W2950525610C2776330181 @default.
• W2950525610 hasConceptScore W2950525610C33923547 @default.
• W2950525610 hasConceptScore W2950525610C41008148 @default.
• W2950525610 hasConceptScore W2950525610C43617362 @default.
• W2950525610 hasConceptScore W2950525610C50644808 @default.
• W2950525610 hasConceptScore W2950525610C51823790 @default.
• W2950525610 hasConceptScore W2950525610C78458016 @default.
• W2950525610 hasConceptScore W2950525610C81917197 @default.

• next