FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2975270761> ?p ?o ?g. }

• W2975270761 abstract "Current deep neural networks can achieve remarkable performance on a single task. However, when the deep neural network is continually trained on a sequence of tasks, it seems to gradually forget the previous learned knowledge. This phenomenon is referred to as catastrophic forgetting and motivates the field called lifelong learning. The central question in lifelong learning is how to enable deep neural networks to maintain performance on old tasks while learning a new this http URL this paper, we introduce a novel and effective lifelong learning algorithm, calledMixEd stochastic GrAdient (MEGA), which allows deep neural networks to ac-quire the ability of retaining performance on old tasks while learning new tasks.MEGA modulates the balance between old tasks and the new task by integrating the current gradient with the gradient computed on a small reference episodic memory. Extensive experimental results show that the proposed MEGA algorithm significantly advances the state-of-the-art on all four commonly used life-long learning benchmarks, reducing the error by up to 18%." @default.
• W2975270761 created "2019-10-03" @default.
• W2975270761 creator A5012033269 @default.
• W2975270761 creator A5023288846 @default.
• W2975270761 creator A5025573294 @default.
• W2975270761 creator A5064437230 @default.
• W2975270761 date "2019-09-25" @default.
• W2975270761 modified "2023-10-11" @default.
• W2975270761 title "Learning with Long-term Remembering: Following the Lead of Mixed Stochastic Gradient." @default.
• W2975270761 cites W1797268635 @default.
• W2975270761 cites W2036963181 @default.
• W2975270761 cites W2060277733 @default.
• W2975270761 cites W2109339818 @default.
• W2975270761 cites W2116522068 @default.
• W2975270761 cites W2134270519 @default.
• W2975270761 cites W2194775991 @default.
• W2975270761 cites W2399033357 @default.
• W2975270761 cites W2432717477 @default.
• W2975270761 cites W2601450892 @default.
• W2975270761 cites W2618767506 @default.
• W2975270761 cites W2734314755 @default.
• W2975270761 cites W2737492962 @default.
• W2975270761 cites W2744654301 @default.
• W2975270761 cites W2765101016 @default.
• W2975270761 cites W2767434619 @default.
• W2975270761 cites W2785567898 @default.
• W2975270761 cites W2786446225 @default.
• W2975270761 cites W2788388592 @default.
• W2975270761 cites W2803552920 @default.
• W2975270761 cites W2804746922 @default.
• W2975270761 cites W2898265486 @default.
• W2975270761 cites W2898290017 @default.
• W2975270761 cites W2912351382 @default.
• W2975270761 cites W2912515466 @default.
• W2975270761 cites W2932484476 @default.
• W2975270761 cites W2949995560 @default.
• W2975270761 cites W2951013084 @default.
• W2975270761 cites W2951544594 @default.
• W2975270761 cites W2951775809 @default.
• W2975270761 cites W2952571470 @default.
• W2975270761 cites W2952677972 @default.
• W2975270761 cites W2963560049 @default.
• W2975270761 cites W2963677766 @default.
• W2975270761 cites W2964790801 @default.
• W2975270761 cites W2967821093 @default.
• W2975270761 cites W2969211464 @default.
• W2975270761 cites W2970586779 @default.
• W2975270761 cites W2980753060 @default.
• W2975270761 cites W2991265870 @default.
• W2975270761 cites W2996653965 @default.
• W2975270761 cites W2997359900 @default.
• W2975270761 cites W3030364939 @default.
• W2975270761 cites W3038082334 @default.
• W2975270761 cites W3118608800 @default.
• W2975270761 cites W2426267443 @default.
• W2975270761 hasPublicationYear "2019" @default.
• W2975270761 type Work @default.
• W2975270761 sameAs 2975270761 @default.
• W2975270761 citedByCount "2" @default.
• W2975270761 countsByYear W29752707612020 @default.
• W2975270761 countsByYear W29752707612021 @default.
• W2975270761 crossrefType "posted-content" @default.
• W2975270761 hasAuthorship W2975270761A5012033269 @default.
• W2975270761 hasAuthorship W2975270761A5023288846 @default.
• W2975270761 hasAuthorship W2975270761A5025573294 @default.
• W2975270761 hasAuthorship W2975270761A5064437230 @default.
• W2975270761 hasConcept C108583219 @default.
• W2975270761 hasConcept C108771440 @default.
• W2975270761 hasConcept C119857082 @default.
• W2975270761 hasConcept C121332964 @default.
• W2975270761 hasConcept C127413603 @default.
• W2975270761 hasConcept C154945302 @default.
• W2975270761 hasConcept C15744967 @default.
• W2975270761 hasConcept C180747234 @default.
• W2975270761 hasConcept C19417346 @default.
• W2975270761 hasConcept C201995342 @default.
• W2975270761 hasConcept C202444582 @default.
• W2975270761 hasConcept C2780451532 @default.
• W2975270761 hasConcept C33923547 @default.
• W2975270761 hasConcept C41008148 @default.
• W2975270761 hasConcept C50644808 @default.
• W2975270761 hasConcept C61797465 @default.
• W2975270761 hasConcept C62520636 @default.
• W2975270761 hasConcept C7149132 @default.
• W2975270761 hasConcept C9652623 @default.
• W2975270761 hasConceptScore W2975270761C108583219 @default.
• W2975270761 hasConceptScore W2975270761C108771440 @default.
• W2975270761 hasConceptScore W2975270761C119857082 @default.
• W2975270761 hasConceptScore W2975270761C121332964 @default.
• W2975270761 hasConceptScore W2975270761C127413603 @default.
• W2975270761 hasConceptScore W2975270761C154945302 @default.
• W2975270761 hasConceptScore W2975270761C15744967 @default.
• W2975270761 hasConceptScore W2975270761C180747234 @default.
• W2975270761 hasConceptScore W2975270761C19417346 @default.
• W2975270761 hasConceptScore W2975270761C201995342 @default.
• W2975270761 hasConceptScore W2975270761C202444582 @default.
• W2975270761 hasConceptScore W2975270761C2780451532 @default.
• W2975270761 hasConceptScore W2975270761C33923547 @default.
• W2975270761 hasConceptScore W2975270761C41008148 @default.
• W2975270761 hasConceptScore W2975270761C50644808 @default.

• next