FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2989753407> ?p ?o ?g. }

• W2989753407 endingPage "11" @default.
• W2989753407 startingPage "1" @default.
• W2989753407 abstract "Optimization algorithms are of great importance to efficiently and effectively train a deep neural network. However, the existing optimization algorithms show unsatisfactory convergence behavior, either slowly converging or not seeking to avoid bad local optima. Learning rate dropout (LRD) is a new gradient descent technique to motivate faster convergence and better generalization. LRD aids the optimizer to actively explore in the parameter space by randomly dropping some learning rates (to 0); at each iteration, only parameters whose learning rate is not 0 are updated. Since LRD reduces the number of parameters to be updated for each iteration, the convergence becomes easier. For parameters that are not updated, their gradients are accumulated (e.g., momentum) by the optimizer for the next update. Accumulating multiple gradients at fixed parameter positions gives the optimizer more energy to escape from the saddle point and bad local optima. Experiments show that LRD is surprisingly effective in accelerating training while preventing overfitting." @default.
• W2989753407 created "2019-12-05" @default.
• W2989753407 creator A5009018575 @default.
• W2989753407 creator A5014165464 @default.
• W2989753407 creator A5040162395 @default.
• W2989753407 creator A5052820597 @default.
• W2989753407 creator A5058405643 @default.
• W2989753407 creator A5085129889 @default.
• W2989753407 date "2022-01-01" @default.
• W2989753407 modified "2023-10-15" @default.
• W2989753407 title "Learning Rate Dropout" @default.
• W2989753407 cites W1522301498 @default.
• W2989753407 cites W1903029394 @default.
• W2989753407 cites W1980287119 @default.
• W2989753407 cites W1994616650 @default.
• W2989753407 cites W2000769684 @default.
• W2989753407 cites W2002745466 @default.
• W2989753407 cites W2031489346 @default.
• W2989753407 cites W2095984592 @default.
• W2989753407 cites W2108677974 @default.
• W2989753407 cites W2112796928 @default.
• W2989753407 cites W2124136621 @default.
• W2989753407 cites W2144794286 @default.
• W2989753407 cites W2145094598 @default.
• W2989753407 cites W2148825261 @default.
• W2989753407 cites W2157444450 @default.
• W2989753407 cites W2194775991 @default.
• W2989753407 cites W2263490141 @default.
• W2989753407 cites W2345474290 @default.
• W2989753407 cites W2504108613 @default.
• W2989753407 cites W2560023338 @default.
• W2989753407 cites W2766736793 @default.
• W2989753407 cites W2785523195 @default.
• W2989753407 cites W2798391154 @default.
• W2989753407 cites W2893810361 @default.
• W2989753407 cites W2907225497 @default.
• W2989753407 cites W2951266961 @default.
• W2989753407 cites W2963208657 @default.
• W2989753407 cites W2963227127 @default.
• W2989753407 cites W2963446712 @default.
• W2989753407 cites W2963563573 @default.
• W2989753407 cites W2963703360 @default.
• W2989753407 cites W2964288706 @default.
• W2989753407 cites W2964309882 @default.
• W2989753407 cites W2968917279 @default.
• W2989753407 cites W3106250896 @default.
• W2989753407 cites W3118608800 @default.
• W2989753407 cites W35527955 @default.
• W2989753407 cites W4919037 @default.
• W2989753407 cites W6908809 @default.
• W2989753407 doi "https://doi.org/10.1109/tnnls.2022.3155181" @default.
• W2989753407 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/35286266" @default.
• W2989753407 hasPublicationYear "2022" @default.
• W2989753407 type Work @default.
• W2989753407 sameAs 2989753407 @default.
• W2989753407 citedByCount "4" @default.
• W2989753407 countsByYear W29897534072020 @default.
• W2989753407 countsByYear W29897534072021 @default.
• W2989753407 countsByYear W29897534072022 @default.
• W2989753407 countsByYear W29897534072023 @default.
• W2989753407 crossrefType "journal-article" @default.
• W2989753407 hasAuthorship W2989753407A5009018575 @default.
• W2989753407 hasAuthorship W2989753407A5014165464 @default.
• W2989753407 hasAuthorship W2989753407A5040162395 @default.
• W2989753407 hasAuthorship W2989753407A5052820597 @default.
• W2989753407 hasAuthorship W2989753407A5058405643 @default.
• W2989753407 hasAuthorship W2989753407A5085129889 @default.
• W2989753407 hasConcept C10138342 @default.
• W2989753407 hasConcept C119857082 @default.
• W2989753407 hasConcept C126255220 @default.
• W2989753407 hasConcept C127162648 @default.
• W2989753407 hasConcept C134306372 @default.
• W2989753407 hasConcept C141934464 @default.
• W2989753407 hasConcept C153258448 @default.
• W2989753407 hasConcept C154945302 @default.
• W2989753407 hasConcept C162324750 @default.
• W2989753407 hasConcept C177148314 @default.
• W2989753407 hasConcept C22019652 @default.
• W2989753407 hasConcept C2524010 @default.
• W2989753407 hasConcept C2681867 @default.
• W2989753407 hasConcept C2776145597 @default.
• W2989753407 hasConcept C2777303404 @default.
• W2989753407 hasConcept C28719098 @default.
• W2989753407 hasConcept C31258907 @default.
• W2989753407 hasConcept C33923547 @default.
• W2989753407 hasConcept C41008148 @default.
• W2989753407 hasConcept C50522688 @default.
• W2989753407 hasConcept C50644808 @default.
• W2989753407 hasConcept C57869625 @default.
• W2989753407 hasConcept C60718061 @default.
• W2989753407 hasConceptScore W2989753407C10138342 @default.
• W2989753407 hasConceptScore W2989753407C119857082 @default.
• W2989753407 hasConceptScore W2989753407C126255220 @default.
• W2989753407 hasConceptScore W2989753407C127162648 @default.
• W2989753407 hasConceptScore W2989753407C134306372 @default.
• W2989753407 hasConceptScore W2989753407C141934464 @default.
• W2989753407 hasConceptScore W2989753407C153258448 @default.
• W2989753407 hasConceptScore W2989753407C154945302 @default.

• next