FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3097629404> ?p ?o ?g. }

• W3097629404 abstract "One of the most fundamental aspects of any machine learning algorithm is the training data used by the algorithm. We introduce the novel concept of $epsilon$-approximation of datasets, obtaining datasets which are much smaller than or are significant corruptions of the original training data while maintaining similar model performance. We introduce a meta-learning algorithm called Kernel Inducing Points (KIP) for obtaining such remarkable datasets, inspired by the recent developments in the correspondence between infinitely-wide neural networks and kernel ridge-regression (KRR). For KRR tasks, we demonstrate that KIP can compress datasets by one or two orders of magnitude, significantly improving previous dataset distillation and subset selection methods while obtaining state of the art results for MNIST and CIFAR-10 classification. Furthermore, our KIP-learned datasets are transferable to the training of finite-width neural networks even beyond the lazy-training regime, which leads to state of the art results for neural network dataset distillation with potential applications to privacy-preservation." @default.
• W3097629404 created "2020-11-09" @default.
• W3097629404 creator A5018500051 @default.
• W3097629404 creator A5064407466 @default.
• W3097629404 creator A5076404171 @default.
• W3097629404 date "2020-10-30" @default.
• W3097629404 modified "2023-09-27" @default.
• W3097629404 title "Dataset Meta-Learning from Kernel Ridge-Regression" @default.
• W3097629404 cites W137285897 @default.
• W3097629404 cites W1605479404 @default.
• W3097629404 cites W2075066642 @default.
• W3097629404 cites W2099768828 @default.
• W3097629404 cites W2112545207 @default.
• W3097629404 cites W2112796928 @default.
• W3097629404 cites W2123872146 @default.
• W3097629404 cites W2135106139 @default.
• W3097629404 cites W2160840682 @default.
• W3097629404 cites W2230030897 @default.
• W3097629404 cites W2473418344 @default.
• W3097629404 cites W2601450892 @default.
• W3097629404 cites W2809090039 @default.
• W3097629404 cites W2910655610 @default.
• W3097629404 cites W2949746412 @default.
• W3097629404 cites W2962685794 @default.
• W3097629404 cites W2963233958 @default.
• W3097629404 cites W2963323437 @default.
• W3097629404 cites W2963341924 @default.
• W3097629404 cites W2964052793 @default.
• W3097629404 cites W2969656782 @default.
• W3097629404 cites W2971043187 @default.
• W3097629404 cites W2980536810 @default.
• W3097629404 cites W2983647115 @default.
• W3097629404 cites W2994747787 @default.
• W3097629404 cites W2994872659 @default.
• W3097629404 cites W3002073503 @default.
• W3097629404 cites W3032949543 @default.
• W3097629404 cites W3035433747 @default.
• W3097629404 cites W3035559424 @default.
• W3097629404 cites W3037120332 @default.
• W3097629404 cites W3101069636 @default.
• W3097629404 cites W3118146262 @default.
• W3097629404 cites W3118608800 @default.
• W3097629404 cites W3119342105 @default.
• W3097629404 hasPublicationYear "2020" @default.
• W3097629404 type Work @default.
• W3097629404 sameAs 3097629404 @default.
• W3097629404 citedByCount "1" @default.
• W3097629404 countsByYear W30976294042021 @default.
• W3097629404 crossrefType "posted-content" @default.
• W3097629404 hasAuthorship W3097629404A5018500051 @default.
• W3097629404 hasAuthorship W3097629404A5064407466 @default.
• W3097629404 hasAuthorship W3097629404A5076404171 @default.
• W3097629404 hasConcept C105795698 @default.
• W3097629404 hasConcept C11413529 @default.
• W3097629404 hasConcept C114614502 @default.
• W3097629404 hasConcept C115903868 @default.
• W3097629404 hasConcept C119857082 @default.
• W3097629404 hasConcept C122280245 @default.
• W3097629404 hasConcept C12267149 @default.
• W3097629404 hasConcept C124101348 @default.
• W3097629404 hasConcept C151730666 @default.
• W3097629404 hasConcept C153180895 @default.
• W3097629404 hasConcept C154945302 @default.
• W3097629404 hasConcept C190502265 @default.
• W3097629404 hasConcept C32277403 @default.
• W3097629404 hasConcept C33923547 @default.
• W3097629404 hasConcept C41008148 @default.
• W3097629404 hasConcept C50644808 @default.
• W3097629404 hasConcept C55166926 @default.
• W3097629404 hasConcept C74193536 @default.
• W3097629404 hasConcept C83546350 @default.
• W3097629404 hasConcept C86803240 @default.
• W3097629404 hasConceptScore W3097629404C105795698 @default.
• W3097629404 hasConceptScore W3097629404C11413529 @default.
• W3097629404 hasConceptScore W3097629404C114614502 @default.
• W3097629404 hasConceptScore W3097629404C115903868 @default.
• W3097629404 hasConceptScore W3097629404C119857082 @default.
• W3097629404 hasConceptScore W3097629404C122280245 @default.
• W3097629404 hasConceptScore W3097629404C12267149 @default.
• W3097629404 hasConceptScore W3097629404C124101348 @default.
• W3097629404 hasConceptScore W3097629404C151730666 @default.
• W3097629404 hasConceptScore W3097629404C153180895 @default.
• W3097629404 hasConceptScore W3097629404C154945302 @default.
• W3097629404 hasConceptScore W3097629404C190502265 @default.
• W3097629404 hasConceptScore W3097629404C32277403 @default.
• W3097629404 hasConceptScore W3097629404C33923547 @default.
• W3097629404 hasConceptScore W3097629404C41008148 @default.
• W3097629404 hasConceptScore W3097629404C50644808 @default.
• W3097629404 hasConceptScore W3097629404C55166926 @default.
• W3097629404 hasConceptScore W3097629404C74193536 @default.
• W3097629404 hasConceptScore W3097629404C83546350 @default.
• W3097629404 hasConceptScore W3097629404C86803240 @default.
• W3097629404 hasLocation W30976294041 @default.
• W3097629404 hasOpenAccess W3097629404 @default.
• W3097629404 hasPrimaryLocation W30976294041 @default.
• W3097629404 hasRelatedWork W1652465939 @default.
• W3097629404 hasRelatedWork W2366165148 @default.
• W3097629404 hasRelatedWork W2611520425 @default.
• W3097629404 hasRelatedWork W2704050264 @default.
• W3097629404 hasRelatedWork W2732561834 @default.

• next