Matches in SemOpenAlex for { <https://semopenalex.org/work/W13488910> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W13488910 endingPage "1289" @default.
- W13488910 startingPage "1283" @default.
- W13488910 abstract "We consider the long-open problem of attribute-efficient learning of halfspaces. In this problem the learner is given random examples labeled by an unknown halfspace functionf on R n . Furtherf is r-sparse, that is it depends on at mostr out ofn variables. An attribute-efficient learning algorithm is an algorithm that can output a hypothesis close to f using a polynomial in r and logn number of examples (Blum, 1992). Despite a number of attempts and some partial progress, there are no efficient algorithms or hardness results for the problem. We propose a potentially easier question: what is the query complexity of this learning problem in the statistical query (SQ) model of Kearns (1998). We show that, as in the case of general PAC learning, the query complexity of attributeefficient SQ learning of any concept class can be characterized by a combinatorial parameter of the concept class. The proposed question is then equivalent to estimating the value of this parameter for the concept class of halfspaces. A potentially simpler problem is to estimate this parameter for the concept class of decision lists, a subclass of halfspaces." @default.
- W13488910 created "2016-06-24" @default.
- W13488910 creator A5054675941 @default.
- W13488910 date "2014-05-29" @default.
- W13488910 modified "2023-09-23" @default.
- W13488910 title "Open Problem: The Statistical Query Complexity of Learning Sparse Halfspaces" @default.
- W13488910 cites W1548189207 @default.
- W13488910 cites W1970472910 @default.
- W13488910 cites W1971718861 @default.
- W13488910 cites W1980379651 @default.
- W13488910 cites W1994022788 @default.
- W13488910 cites W1995897489 @default.
- W13488910 cites W2023146810 @default.
- W13488910 cites W2040884411 @default.
- W13488910 cites W2059602918 @default.
- W13488910 cites W2064680241 @default.
- W13488910 cites W2097886040 @default.
- W13488910 cites W2098730982 @default.
- W13488910 cites W2114522915 @default.
- W13488910 cites W2122537655 @default.
- W13488910 cites W2126177819 @default.
- W13488910 cites W2129113961 @default.
- W13488910 cites W2130608487 @default.
- W13488910 cites W2151654130 @default.
- W13488910 cites W2155831576 @default.
- W13488910 cites W2163361804 @default.
- W13488910 cites W2187362426 @default.
- W13488910 cites W2546502841 @default.
- W13488910 cites W2912052665 @default.
- W13488910 cites W2964189588 @default.
- W13488910 cites W2964247925 @default.
- W13488910 cites W2964345707 @default.
- W13488910 cites W3023412376 @default.
- W13488910 cites W58673994 @default.
- W13488910 hasPublicationYear "2014" @default.
- W13488910 type Work @default.
- W13488910 sameAs 13488910 @default.
- W13488910 citedByCount "5" @default.
- W13488910 countsByYear W134889102014 @default.
- W13488910 countsByYear W134889102015 @default.
- W13488910 countsByYear W134889102016 @default.
- W13488910 countsByYear W134889102020 @default.
- W13488910 crossrefType "proceedings-article" @default.
- W13488910 hasAuthorship W13488910A5054675941 @default.
- W13488910 hasConcept C11413529 @default.
- W13488910 hasConcept C114614502 @default.
- W13488910 hasConcept C154945302 @default.
- W13488910 hasConcept C179799912 @default.
- W13488910 hasConcept C203313322 @default.
- W13488910 hasConcept C2777212361 @default.
- W13488910 hasConcept C3018263672 @default.
- W13488910 hasConcept C311688 @default.
- W13488910 hasConcept C33923547 @default.
- W13488910 hasConcept C41008148 @default.
- W13488910 hasConcept C50292564 @default.
- W13488910 hasConcept C77967617 @default.
- W13488910 hasConcept C80444323 @default.
- W13488910 hasConceptScore W13488910C11413529 @default.
- W13488910 hasConceptScore W13488910C114614502 @default.
- W13488910 hasConceptScore W13488910C154945302 @default.
- W13488910 hasConceptScore W13488910C179799912 @default.
- W13488910 hasConceptScore W13488910C203313322 @default.
- W13488910 hasConceptScore W13488910C2777212361 @default.
- W13488910 hasConceptScore W13488910C3018263672 @default.
- W13488910 hasConceptScore W13488910C311688 @default.
- W13488910 hasConceptScore W13488910C33923547 @default.
- W13488910 hasConceptScore W13488910C41008148 @default.
- W13488910 hasConceptScore W13488910C50292564 @default.
- W13488910 hasConceptScore W13488910C77967617 @default.
- W13488910 hasConceptScore W13488910C80444323 @default.
- W13488910 hasOpenAccess W13488910 @default.
- W13488910 hasRelatedWork W1980379651 @default.
- W13488910 hasRelatedWork W1995897489 @default.
- W13488910 hasRelatedWork W2006412101 @default.
- W13488910 hasRelatedWork W2010523825 @default.
- W13488910 hasRelatedWork W2019363670 @default.
- W13488910 hasRelatedWork W2020007442 @default.
- W13488910 hasRelatedWork W2064680241 @default.
- W13488910 hasRelatedWork W2109722477 @default.
- W13488910 hasRelatedWork W2149840272 @default.
- W13488910 hasRelatedWork W2155831576 @default.
- W13488910 hasRelatedWork W2186076924 @default.
- W13488910 hasRelatedWork W2201177080 @default.
- W13488910 hasRelatedWork W2245160765 @default.
- W13488910 hasRelatedWork W2401071017 @default.
- W13488910 hasRelatedWork W2546502841 @default.
- W13488910 hasRelatedWork W2964276008 @default.
- W13488910 hasRelatedWork W3033838839 @default.
- W13488910 hasRelatedWork W3117722141 @default.
- W13488910 hasRelatedWork W3034104366 @default.
- W13488910 hasRelatedWork W3184572569 @default.
- W13488910 isParatext "false" @default.
- W13488910 isRetracted "false" @default.
- W13488910 magId "13488910" @default.
- W13488910 workType "article" @default.