Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897032526> ?p ?o ?g. }
Showing items 1 to 84 of
84
with 100 items per page.
- W2897032526 endingPage "746" @default.
- W2897032526 startingPage "727" @default.
- W2897032526 abstract "We show that for the problem of testing if a matrix A ∈ Fn × n has rank at most d, or requires changing an ϵ-fraction of entries to have rank at most d, there is a non-adaptive query algorithm making O(d2/ϵ) queries. Our algorithm works for any field F. This improves upon the previous O(d2/ϵ2) bound (Krauthgamer and Sasson, SODA '03), and bypasses an Ω(d2/ϵ2) lower bound of (Li, Wang, and Woodruff, KDD '14) which holds if the algorithm is required to read a submatrix. Our algorithm is the first such algorithm which does not read a submatrix, and instead reads a carefully selected non-adaptive pattern of entries in rows and columns of A. We complement our algorithm with a matching [MATH HERE] query complexity lower bound for non-adaptive testers over any field. We also give tight bounds of [MATH HERE] queries in the sensing model for which query access comes in the form of 〈Xi, A〉 ≔ tr(Xi⊤A); perhaps surprisingly these bounds do not depend on ϵ.Testing rank is only one of many tasks in determining if a matrix has low intrinsic dimensionality We next develop a novel property testing framework for testing numerical properties of a real-valued matrix A more generally, which includes the stable rank, Schatten-p norms, and SVD entropy. Specifically, we propose a bounded entry model, where A is required to have entries bounded by 1 in absolute value. Such a model provides a meaningful framework for testing numerical quantities and avoids trivialities caused by single entries being arbitrarily large. It is also well-motivated by recommendation systems. We give upper and lower bounds for a wide range of problems in this model, and discuss connections to the sensing model above. We obtain several results for estimating the operator norm that may be of independent interest. For example, we show that if the stable rank is constant, ||A||F = Ω(n), and the singular value gap σ1(A)/σ2(A) = (1/ϵ)γ for any constant γ > 0, then the operator norm can be estimated up to a (1 ± ϵ)-factor non-adaptively by querying O(1/ϵ2) entries. This should be contrasted to adaptive methods such as the power method, or previous non-adaptive sampling schemes based on matrix Bernstein inequalities which read a 1/ϵ2 × 1/ϵ2 submatrix and thus make Ω(1/ϵ4) queries. Similar to our non-adaptive algorithm for testing rank, our scheme instead reads a carefully selected pattern of entries." @default.
- W2897032526 created "2018-10-26" @default.
- W2897032526 creator A5018348928 @default.
- W2897032526 creator A5024805876 @default.
- W2897032526 creator A5064842058 @default.
- W2897032526 creator A5068544954 @default.
- W2897032526 date "2019-01-06" @default.
- W2897032526 modified "2023-09-27" @default.
- W2897032526 title "Testing matrix rank, optimally" @default.
- W2897032526 doi "https://doi.org/10.5555/3310435.3310481" @default.
- W2897032526 hasPublicationYear "2019" @default.
- W2897032526 type Work @default.
- W2897032526 sameAs 2897032526 @default.
- W2897032526 citedByCount "10" @default.
- W2897032526 countsByYear W28970325262019 @default.
- W2897032526 countsByYear W28970325262020 @default.
- W2897032526 countsByYear W28970325262021 @default.
- W2897032526 countsByYear W28970325262022 @default.
- W2897032526 crossrefType "proceedings-article" @default.
- W2897032526 hasAuthorship W2897032526A5018348928 @default.
- W2897032526 hasAuthorship W2897032526A5024805876 @default.
- W2897032526 hasAuthorship W2897032526A5064842058 @default.
- W2897032526 hasAuthorship W2897032526A5068544954 @default.
- W2897032526 hasConcept C105795698 @default.
- W2897032526 hasConcept C106487976 @default.
- W2897032526 hasConcept C11413529 @default.
- W2897032526 hasConcept C114614502 @default.
- W2897032526 hasConcept C118615104 @default.
- W2897032526 hasConcept C123842658 @default.
- W2897032526 hasConcept C134306372 @default.
- W2897032526 hasConcept C159985019 @default.
- W2897032526 hasConcept C164226766 @default.
- W2897032526 hasConcept C165064840 @default.
- W2897032526 hasConcept C192562407 @default.
- W2897032526 hasConcept C22789450 @default.
- W2897032526 hasConcept C33923547 @default.
- W2897032526 hasConcept C34388435 @default.
- W2897032526 hasConcept C41008148 @default.
- W2897032526 hasConcept C77553402 @default.
- W2897032526 hasConceptScore W2897032526C105795698 @default.
- W2897032526 hasConceptScore W2897032526C106487976 @default.
- W2897032526 hasConceptScore W2897032526C11413529 @default.
- W2897032526 hasConceptScore W2897032526C114614502 @default.
- W2897032526 hasConceptScore W2897032526C118615104 @default.
- W2897032526 hasConceptScore W2897032526C123842658 @default.
- W2897032526 hasConceptScore W2897032526C134306372 @default.
- W2897032526 hasConceptScore W2897032526C159985019 @default.
- W2897032526 hasConceptScore W2897032526C164226766 @default.
- W2897032526 hasConceptScore W2897032526C165064840 @default.
- W2897032526 hasConceptScore W2897032526C192562407 @default.
- W2897032526 hasConceptScore W2897032526C22789450 @default.
- W2897032526 hasConceptScore W2897032526C33923547 @default.
- W2897032526 hasConceptScore W2897032526C34388435 @default.
- W2897032526 hasConceptScore W2897032526C41008148 @default.
- W2897032526 hasConceptScore W2897032526C77553402 @default.
- W2897032526 hasLocation W28970325261 @default.
- W2897032526 hasOpenAccess W2897032526 @default.
- W2897032526 hasPrimaryLocation W28970325261 @default.
- W2897032526 hasRelatedWork W1583642434 @default.
- W2897032526 hasRelatedWork W1770971269 @default.
- W2897032526 hasRelatedWork W1983068660 @default.
- W2897032526 hasRelatedWork W2064868049 @default.
- W2897032526 hasRelatedWork W2075290628 @default.
- W2897032526 hasRelatedWork W2095030632 @default.
- W2897032526 hasRelatedWork W2293297835 @default.
- W2897032526 hasRelatedWork W2517080291 @default.
- W2897032526 hasRelatedWork W2754879912 @default.
- W2897032526 hasRelatedWork W2778880769 @default.
- W2897032526 hasRelatedWork W2788387635 @default.
- W2897032526 hasRelatedWork W2914377170 @default.
- W2897032526 hasRelatedWork W2949947345 @default.
- W2897032526 hasRelatedWork W2964075618 @default.
- W2897032526 hasRelatedWork W2996763839 @default.
- W2897032526 hasRelatedWork W3021409470 @default.
- W2897032526 hasRelatedWork W3022155794 @default.
- W2897032526 hasRelatedWork W3022305591 @default.
- W2897032526 hasRelatedWork W3161903763 @default.
- W2897032526 hasRelatedWork W3002604948 @default.
- W2897032526 isParatext "false" @default.
- W2897032526 isRetracted "false" @default.
- W2897032526 magId "2897032526" @default.
- W2897032526 workType "article" @default.