Least squares approximations of measures via geometric condition numbers
From MaRDI portal
Publication:3115487
Abstract: For a probability measure on a real separable Hilbert space, we are interested in "volume-based" approximations of the d-dimensional least squares error of it, i.e., least squares error with respect to a best fit d-dimensional affine subspace. Such approximations are given by averaging real-valued multivariate functions which are typically scalings of squared (d+1)-volumes of (d+1)-simplices. Specifically, we show that such averages are comparable to the square of the d-dimensional least squares error of that measure, where the comparison depends on a simple quantitative geometric property of it. This result is a higher dimensional generalization of the elementary fact that the double integral of the squared distances between points is proportional to the variance of measure. We relate our work to two recent algorithms, one for clustering affine subspaces and the other for Monte-Carlo SVD based on volume sampling.
Recommendations
- On the stability and accuracy of least squares approximations
- High-dimensional Menger-type curvatures. II: \(d\)-separation and a menagerie of curvatures
- Optimal weighted least-squares methods
- Constructing least-squares polynomial approximations
- On multiplicative \(\lambda\)-approximations and some geometric applications
Cites work
- Adaptive Sampling and Fast Low-Rank Matrix Approximation
- Foundations of a multi-way spectral clustering framework for hybrid linear modeling
- High-dimensional Menger-type curvatures. I: Geometric multipoles and multiscale inequalities
- High-dimensional Menger-type curvatures. II: \(d\)-separation and a menagerie of curvatures
- Integral Menger curvature for surfaces
- Matrix approximation and projective clustering via volume sampling
- Menger curvature and rectifiability
- On \(d\)-dimensional \(d\)-semimetrics and simplex-type inequalities for high-dimensional sine functions
- Principal values for Riesz transforms and rectifiability
- Spectral clustering based on local linear approximations
- The Cauchy integral, analytic capacity, and uniform rectifiability
Cited in
(7)- The Condition Number of Riemannian Approximation Problems
- Characterizations of countably \(n\)-rectifiable Radon measures by higher-dimensional Menger curvatures
- Least squares and approximate equidistribution in multidimensions
- An analyst's traveling salesman theorem for sets of dimension larger than one
- High-dimensional Menger-type curvatures. I: Geometric multipoles and multiscale inequalities
- Menger curvatures and \(C^{1,\alpha}\) rectifiability of measures
- High-dimensional Menger-type curvatures. II: \(d\)-separation and a menagerie of curvatures
This page was built for publication: Least squares approximations of measures via geometric condition numbers
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3115487)
