Matches in SemOpenAlex for { <https://semopenalex.org/work/W2913773024> ?p ?o ?g. }
- W2913773024 abstract "The performance of policy gradient methods is sensitive to hyperparameter settings that must be tuned for any new application. Widely used grid search methods for tuning hyperparameters are sample inefficient and computationally expensive. More advanced methods like Population Based Training that learn optimal schedules for hyperparameters instead of fixed settings can yield better results, but are also sample inefficient and computationally expensive. In this paper, we propose Hyperparameter Optimisation on the Fly (HOOF), a gradient-free algorithm that requires no more than one training run to automatically adapt the hyperparameter that affect the policy update directly through the gradient. The main idea is to use existing trajectories sampled by the policy gradient method to optimise a one-step improvement objective, yielding a sample and computationally efficient algorithm that is easy to implement. Our experimental results across multiple domains and algorithms show that using HOOF to learn these hyperparameter schedules leads to faster learning with improved performance." @default.
- W2913773024 created "2019-02-21" @default.
- W2913773024 creator A5030103572 @default.
- W2913773024 creator A5043731054 @default.
- W2913773024 creator A5056879203 @default.
- W2913773024 date "2019-04-27" @default.
- W2913773024 modified "2023-09-27" @default.
- W2913773024 title "Fast Efficient Hyperparameter Tuning for Policy Gradients" @default.
- W2913773024 cites W1533803232 @default.
- W2913773024 cites W1569788011 @default.
- W2913773024 cites W1583330603 @default.
- W2913773024 cites W1771410628 @default.
- W2913773024 cites W2097998348 @default.
- W2913773024 cites W2099201756 @default.
- W2913773024 cites W2102863375 @default.
- W2913773024 cites W2106411961 @default.
- W2913773024 cites W2112036188 @default.
- W2913773024 cites W2113145584 @default.
- W2913773024 cites W2119076496 @default.
- W2913773024 cites W2119717200 @default.
- W2913773024 cites W2130801532 @default.
- W2913773024 cites W2131241448 @default.
- W2913773024 cites W2155027007 @default.
- W2913773024 cites W2166107799 @default.
- W2913773024 cites W2175583736 @default.
- W2913773024 cites W2186601217 @default.
- W2913773024 cites W2194966727 @default.
- W2913773024 cites W2270245739 @default.
- W2913773024 cites W2277643282 @default.
- W2913773024 cites W2408019865 @default.
- W2913773024 cites W2434014514 @default.
- W2913773024 cites W2604763608 @default.
- W2913773024 cites W2616619952 @default.
- W2913773024 cites W2624731731 @default.
- W2913773024 cites W2788049526 @default.
- W2913773024 cites W2799181865 @default.
- W2913773024 cites W2890803796 @default.
- W2913773024 cites W2895212759 @default.
- W2913773024 cites W2903717821 @default.
- W2913773024 cites W2916338919 @default.
- W2913773024 cites W2949608212 @default.
- W2913773024 cites W2951665052 @default.
- W2913773024 cites W2962749646 @default.
- W2913773024 cites W2962779867 @default.
- W2913773024 cites W2962854145 @default.
- W2913773024 cites W2963120839 @default.
- W2913773024 cites W2963184621 @default.
- W2913773024 cites W2963394426 @default.
- W2913773024 cites W2963641140 @default.
- W2913773024 cites W2964029277 @default.
- W2913773024 cites W2964043796 @default.
- W2913773024 cites W2964121744 @default.
- W2913773024 cites W2964296021 @default.
- W2913773024 cites W3037207827 @default.
- W2913773024 cites W3102150134 @default.
- W2913773024 cites W3136620014 @default.
- W2913773024 cites W567721252 @default.
- W2913773024 cites W60686164 @default.
- W2913773024 cites W2770298516 @default.
- W2913773024 hasPublicationYear "2019" @default.
- W2913773024 type Work @default.
- W2913773024 sameAs 2913773024 @default.
- W2913773024 citedByCount "9" @default.
- W2913773024 countsByYear W29137730242019 @default.
- W2913773024 countsByYear W29137730242020 @default.
- W2913773024 countsByYear W29137730242021 @default.
- W2913773024 crossrefType "posted-content" @default.
- W2913773024 hasAuthorship W2913773024A5030103572 @default.
- W2913773024 hasAuthorship W2913773024A5043731054 @default.
- W2913773024 hasAuthorship W2913773024A5056879203 @default.
- W2913773024 hasConcept C10485038 @default.
- W2913773024 hasConcept C11413529 @default.
- W2913773024 hasConcept C119857082 @default.
- W2913773024 hasConcept C12267149 @default.
- W2913773024 hasConcept C126255220 @default.
- W2913773024 hasConcept C154945302 @default.
- W2913773024 hasConcept C185592680 @default.
- W2913773024 hasConcept C187691185 @default.
- W2913773024 hasConcept C198531522 @default.
- W2913773024 hasConcept C2524010 @default.
- W2913773024 hasConcept C33923547 @default.
- W2913773024 hasConcept C41008148 @default.
- W2913773024 hasConcept C43617362 @default.
- W2913773024 hasConcept C8642999 @default.
- W2913773024 hasConceptScore W2913773024C10485038 @default.
- W2913773024 hasConceptScore W2913773024C11413529 @default.
- W2913773024 hasConceptScore W2913773024C119857082 @default.
- W2913773024 hasConceptScore W2913773024C12267149 @default.
- W2913773024 hasConceptScore W2913773024C126255220 @default.
- W2913773024 hasConceptScore W2913773024C154945302 @default.
- W2913773024 hasConceptScore W2913773024C185592680 @default.
- W2913773024 hasConceptScore W2913773024C187691185 @default.
- W2913773024 hasConceptScore W2913773024C198531522 @default.
- W2913773024 hasConceptScore W2913773024C2524010 @default.
- W2913773024 hasConceptScore W2913773024C33923547 @default.
- W2913773024 hasConceptScore W2913773024C41008148 @default.
- W2913773024 hasConceptScore W2913773024C43617362 @default.
- W2913773024 hasConceptScore W2913773024C8642999 @default.
- W2913773024 hasLocation W29137730241 @default.
- W2913773024 hasOpenAccess W2913773024 @default.