Matches in SemOpenAlex for { <https://semopenalex.org/work/W2743151379> ?p ?o ?g. }
- W2743151379 abstract "Deep neural networks are used in many state-of-the-art systems for machine perception. Once a network is trained to do a specific task, e.g., bird classification, it cannot easily be trained to do new tasks, e.g., incrementally learning to recognize additional bird species or learning an entirely different task such as flower recognition. When new tasks are added, typical deep neural networks are prone to catastrophically forgetting previous tasks. Networks that are capable of assimilating new information incrementally, much like how humans form new memories over time, will be more efficient than re-training the model from scratch each time a new task needs to be learned. There have been multiple attempts to develop schemes that mitigate catastrophic forgetting, but these methods have not been directly compared, the tests used to evaluate them vary considerably, and these methods have only been evaluated on small-scale problems (e.g., MNIST). In this paper, we introduce new metrics and benchmarks for directly comparing five different mechanisms designed to mitigate catastrophic forgetting in neural networks: regularization, ensembling, rehearsal, dual-memory, and sparse-coding. Our experiments on real-world images and sounds show that the mechanism(s) that are critical for optimal performance vary based on the incremental training paradigm and type of data being used, but they all demonstrate that the catastrophic forgetting problem is not yet solved." @default.
- W2743151379 created "2017-08-17" @default.
- W2743151379 creator A5007265221 @default.
- W2743151379 creator A5030235071 @default.
- W2743151379 creator A5030503283 @default.
- W2743151379 creator A5043035821 @default.
- W2743151379 creator A5046979072 @default.
- W2743151379 date "2018-04-29" @default.
- W2743151379 modified "2023-10-14" @default.
- W2743151379 title "Measuring Catastrophic Forgetting in Neural Networks" @default.
- W2743151379 cites W1525859397 @default.
- W2743151379 cites W1573706465 @default.
- W2743151379 cites W1757796397 @default.
- W2743151379 cites W1797268635 @default.
- W2743151379 cites W1903029394 @default.
- W2743151379 cites W1977544285 @default.
- W2743151379 cites W1996340400 @default.
- W2743151379 cites W2018124860 @default.
- W2743151379 cites W2103753221 @default.
- W2743151379 cites W2113839990 @default.
- W2743151379 cites W2116522068 @default.
- W2743151379 cites W2117539524 @default.
- W2743151379 cites W2122838776 @default.
- W2743151379 cites W2143386621 @default.
- W2743151379 cites W2194775991 @default.
- W2743151379 cites W2401932923 @default.
- W2743151379 cites W2524365899 @default.
- W2743151379 cites W2560647685 @default.
- W2743151379 cites W2564752792 @default.
- W2743151379 cites W2583761661 @default.
- W2743151379 cites W2593116425 @default.
- W2743151379 cites W2605353576 @default.
- W2743151379 cites W3103800629 @default.
- W2743151379 doi "https://doi.org/10.1609/aaai.v32i1.11651" @default.
- W2743151379 hasPublicationYear "2018" @default.
- W2743151379 type Work @default.
- W2743151379 sameAs 2743151379 @default.
- W2743151379 citedByCount "194" @default.
- W2743151379 countsByYear W27431513792017 @default.
- W2743151379 countsByYear W27431513792018 @default.
- W2743151379 countsByYear W27431513792019 @default.
- W2743151379 countsByYear W27431513792020 @default.
- W2743151379 countsByYear W27431513792021 @default.
- W2743151379 countsByYear W27431513792022 @default.
- W2743151379 countsByYear W27431513792023 @default.
- W2743151379 crossrefType "journal-article" @default.
- W2743151379 hasAuthorship W2743151379A5007265221 @default.
- W2743151379 hasAuthorship W2743151379A5030235071 @default.
- W2743151379 hasAuthorship W2743151379A5030503283 @default.
- W2743151379 hasAuthorship W2743151379A5043035821 @default.
- W2743151379 hasAuthorship W2743151379A5046979072 @default.
- W2743151379 hasBestOaLocation W27431513791 @default.
- W2743151379 hasConcept C108583219 @default.
- W2743151379 hasConcept C119857082 @default.
- W2743151379 hasConcept C127413603 @default.
- W2743151379 hasConcept C138885662 @default.
- W2743151379 hasConcept C154945302 @default.
- W2743151379 hasConcept C169760540 @default.
- W2743151379 hasConcept C190502265 @default.
- W2743151379 hasConcept C201995342 @default.
- W2743151379 hasConcept C26760741 @default.
- W2743151379 hasConcept C2780451532 @default.
- W2743151379 hasConcept C2984842247 @default.
- W2743151379 hasConcept C41008148 @default.
- W2743151379 hasConcept C41895202 @default.
- W2743151379 hasConcept C50644808 @default.
- W2743151379 hasConcept C7149132 @default.
- W2743151379 hasConcept C86803240 @default.
- W2743151379 hasConceptScore W2743151379C108583219 @default.
- W2743151379 hasConceptScore W2743151379C119857082 @default.
- W2743151379 hasConceptScore W2743151379C127413603 @default.
- W2743151379 hasConceptScore W2743151379C138885662 @default.
- W2743151379 hasConceptScore W2743151379C154945302 @default.
- W2743151379 hasConceptScore W2743151379C169760540 @default.
- W2743151379 hasConceptScore W2743151379C190502265 @default.
- W2743151379 hasConceptScore W2743151379C201995342 @default.
- W2743151379 hasConceptScore W2743151379C26760741 @default.
- W2743151379 hasConceptScore W2743151379C2780451532 @default.
- W2743151379 hasConceptScore W2743151379C2984842247 @default.
- W2743151379 hasConceptScore W2743151379C41008148 @default.
- W2743151379 hasConceptScore W2743151379C41895202 @default.
- W2743151379 hasConceptScore W2743151379C50644808 @default.
- W2743151379 hasConceptScore W2743151379C7149132 @default.
- W2743151379 hasConceptScore W2743151379C86803240 @default.
- W2743151379 hasIssue "1" @default.
- W2743151379 hasLocation W27431513791 @default.
- W2743151379 hasLocation W27431513792 @default.
- W2743151379 hasOpenAccess W2743151379 @default.
- W2743151379 hasPrimaryLocation W27431513791 @default.
- W2743151379 hasRelatedWork W2618574054 @default.
- W2743151379 hasRelatedWork W2799384463 @default.
- W2743151379 hasRelatedWork W2989784533 @default.
- W2743151379 hasRelatedWork W2996058201 @default.
- W2743151379 hasRelatedWork W3018979822 @default.
- W2743151379 hasRelatedWork W3026616975 @default.
- W2743151379 hasRelatedWork W4288018014 @default.
- W2743151379 hasRelatedWork W4297776111 @default.
- W2743151379 hasRelatedWork W4377865163 @default.
- W2743151379 hasRelatedWork W4385524141 @default.
- W2743151379 hasVolume "32" @default.