Matches in SemOpenAlex for { <https://semopenalex.org/work/W2901064314> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W2901064314 abstract "Catastrophic forgetting is a tough issue when the agent faces the sequential multi-task learning scenario without storing previous task information. It gradually becomes an obstacle to achieve artificial general intelligence which is generally believed to behave like a human with continuous learning capability. In this paper, we propose to utilize the variational Bayesian inference method to overcome catastrophic forgetting. By pruning the neural network according to the mean and variance of weights, parameters are vastly reduced, which mitigates the storage problem of double parameters required in variational Bayesian inference. Based on this lightweight version, autoencoders trained on different tasks are employed to self-adaptively match the corresponding task parameters to tackle sequential multi-task learning problem. We show experimentally on several fundamental datasets that the proposed method can perform substantial improvements without catastrophic forgetting over other classic methods especially in the setting where the probability distributions between tasks present more different." @default.
- W2901064314 created "2018-11-29" @default.
- W2901064314 creator A5007963696 @default.
- W2901064314 creator A5023214008 @default.
- W2901064314 creator A5056039500 @default.
- W2901064314 date "2018-01-01" @default.
- W2901064314 modified "2023-09-25" @default.
- W2901064314 title "Overcoming Catastrophic Forgetting with Self-adaptive Identifiers" @default.
- W2901064314 cites W2102605133 @default.
- W2901064314 cites W2126204609 @default.
- W2901064314 cites W2473930607 @default.
- W2901064314 cites W2560647685 @default.
- W2901064314 cites W2791869553 @default.
- W2901064314 doi "https://doi.org/10.1007/978-3-030-04182-3_43" @default.
- W2901064314 hasPublicationYear "2018" @default.
- W2901064314 type Work @default.
- W2901064314 sameAs 2901064314 @default.
- W2901064314 citedByCount "0" @default.
- W2901064314 crossrefType "book-chapter" @default.
- W2901064314 hasAuthorship W2901064314A5007963696 @default.
- W2901064314 hasAuthorship W2901064314A5023214008 @default.
- W2901064314 hasAuthorship W2901064314A5056039500 @default.
- W2901064314 hasConcept C107673813 @default.
- W2901064314 hasConcept C108010975 @default.
- W2901064314 hasConcept C119857082 @default.
- W2901064314 hasConcept C138885662 @default.
- W2901064314 hasConcept C154945302 @default.
- W2901064314 hasConcept C162324750 @default.
- W2901064314 hasConcept C187736073 @default.
- W2901064314 hasConcept C2776214188 @default.
- W2901064314 hasConcept C2780451532 @default.
- W2901064314 hasConcept C41008148 @default.
- W2901064314 hasConcept C41895202 @default.
- W2901064314 hasConcept C50644808 @default.
- W2901064314 hasConcept C6557445 @default.
- W2901064314 hasConcept C7149132 @default.
- W2901064314 hasConcept C86803240 @default.
- W2901064314 hasConceptScore W2901064314C107673813 @default.
- W2901064314 hasConceptScore W2901064314C108010975 @default.
- W2901064314 hasConceptScore W2901064314C119857082 @default.
- W2901064314 hasConceptScore W2901064314C138885662 @default.
- W2901064314 hasConceptScore W2901064314C154945302 @default.
- W2901064314 hasConceptScore W2901064314C162324750 @default.
- W2901064314 hasConceptScore W2901064314C187736073 @default.
- W2901064314 hasConceptScore W2901064314C2776214188 @default.
- W2901064314 hasConceptScore W2901064314C2780451532 @default.
- W2901064314 hasConceptScore W2901064314C41008148 @default.
- W2901064314 hasConceptScore W2901064314C41895202 @default.
- W2901064314 hasConceptScore W2901064314C50644808 @default.
- W2901064314 hasConceptScore W2901064314C6557445 @default.
- W2901064314 hasConceptScore W2901064314C7149132 @default.
- W2901064314 hasConceptScore W2901064314C86803240 @default.
- W2901064314 hasLocation W29010643141 @default.
- W2901064314 hasOpenAccess W2901064314 @default.
- W2901064314 hasPrimaryLocation W29010643141 @default.
- W2901064314 hasRelatedWork W2474280151 @default.
- W2901064314 hasRelatedWork W2809558795 @default.
- W2901064314 hasRelatedWork W2954466931 @default.
- W2901064314 hasRelatedWork W2955830853 @default.
- W2901064314 hasRelatedWork W2962736944 @default.
- W2901064314 hasRelatedWork W2972768329 @default.
- W2901064314 hasRelatedWork W2994887008 @default.
- W2901064314 hasRelatedWork W2996486651 @default.
- W2901064314 hasRelatedWork W2996653965 @default.
- W2901064314 hasRelatedWork W3000074789 @default.
- W2901064314 hasRelatedWork W3022059646 @default.
- W2901064314 hasRelatedWork W3034856281 @default.
- W2901064314 hasRelatedWork W3036362369 @default.
- W2901064314 hasRelatedWork W3082871683 @default.
- W2901064314 hasRelatedWork W3112858938 @default.
- W2901064314 hasRelatedWork W3114102820 @default.
- W2901064314 hasRelatedWork W3128395934 @default.
- W2901064314 hasRelatedWork W3174716214 @default.
- W2901064314 hasRelatedWork W3190534097 @default.
- W2901064314 hasRelatedWork W3197733496 @default.
- W2901064314 isParatext "false" @default.
- W2901064314 isRetracted "false" @default.
- W2901064314 magId "2901064314" @default.
- W2901064314 workType "book-chapter" @default.