A new look at random projections of the cube and general product measures

DOI10.3150/20-BEJ1303MaRDI QIDQ2040109zbMATH OpenOpenAlexFDO

Authors Zakhar Kabluchko, Joscha Prochno, Christoph Thäle

Publication date 9 July 2021

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1910.02676

zbMATH Keywords

cubeGaussian projectionsGaussian random matricesHausdorff distancehigh-dimensional probabilitylarge deviationslaw of large numbersrandom projectionsStiefel manifold

Mathematics Subject Classification ID

Geometric probability and stochastic geometry (60D05) Large deviations (60F10) Strong limit theorems (60F15)

Abstract: A strong law of large numbers for $d$-dimensional random projections of the $n$-dimensional cube is derived. It shows that with respect to the Hausdorff distance a properly normalized random projection of $[-1,1{]}^n$ onto $mathbb{R}^d$ almost surely converges to a centered $d$-dimensional Euclidean ball of radius $sqrt2/pi$, as $noinfty$. For every point inside this ball we determine the asymptotic number of vertices and the volume of the part of the cube projected `close' to this point. Moreover, large deviations for random projections of general product measures are studied. Let $u^{otimesn}$ be the $n$-fold product measure of a Borel probability measure $u$ on $mathbbR$, and let $I$ be uniformly distributed on the Stiefel manifold of orthogonal $d$-frames in $mathbb{R}^n$. It is shown that the sequence of random measures $u^{otimesn}circ(n^{-1/2}{I}^{\ast }{)}^{-1}$, $ninmathbbN$, satisfies a large deviations principle with probability $1$. The rate function is explicitly identified in terms of the moment generating function of $u$. At the heart of the proofs lies a transition trick which allows to replace the uniform projection by the Gaussian one. A number of concrete examples are discussed as well, including the uniform distributions on the cube $[-1,1{]}^n$ and the discrete cube ${-1,1}^n$ as a special cases.

Review text

The authors indicate that the Dvoretzky-Milman theorem implies a strong law of large numbers for \(d\)-dimensional random projections of the \(n\)-dimensional cube. It is shown that, with respect to the Hausdorff distance, a uniform random projection of the cube \({\left[ { - {1 \mathord{\left/ {\vphantom {1 {\sqrt n }}} \right. \kern-\nulldelimiterspace} {\sqrt n }},{1 \mathord{\left/ {\vphantom {1 {\sqrt n }}} \right. \kern-\nulldelimiterspace} {\sqrt n }}} \right]^n}\) onto \({\mathbb R^d}\) converges almost surely to a centered \(d\)-dimensional Euclidean ball of radius \(\sqrt {{2 \mathord{\left/ {\vphantom {2 \pi }} \right. \kern-\nulldelimiterspace} \pi }} \), as \(n \to \infty \). The paper starts by providing an alternative proof of this strong law via the Artstein-Vitale law of large numbers for random compact sets. Next, it gives an explicit description of the large deviations of an \(n\)-fold product measure under \(d\)-dimensional random projections, where \(d\) is fixed and \(n \to \infty \).

Recommendations

• scientific article; zbMATH DE number 3944989
• A central limit theorem for projections of the cube
• The variance conjecture on projections of the cube
• Projections of the uniform distribution on the cube: a large deviation perspective
• Product of random projections, Jacobi ensembles and universality problems arising from free probability
• On the random product of orthogonal projections in Hilbert space
• \(L^{q}\) dimensions and projections of random measures
• Non-degenerate Hilbert cubes in random sets
• Projections of random Cantor sets
• On the random product of orthogonal projections in Hilbert space. II

Cites work

• 4. On the power of random information
• A central limit theorem for convex sets
• A central limit theorem for projections of the cube
• A conditional limit theorem for high-dimensional ^p-spheres
• A strong law of large numbers for random compact sets
• Alice and Bob Meet Banach
• Asymptotic geometric analysis. I
• Asymptotic theory of finite dimensional normed spaces. With an appendix by M. Gromov: Isoperimetric inequalities in Riemannian manifolds
• CLT and the volume of intersections of \(\ell_p^n\)-balls
• Counting the faces of randomly-projected hypercubes and orthants, with applications
• Entropy numbers of embeddings of Schatten classes
• Gaussian fluctuations for high-dimensional random projections of \(\ell_p^n\)-balls
• Geometry of _p^n-balls: classical results and recent developments
• Geometry of families of random projections of symmetric convex bodies.
• High-dimensional limit theorems for random vectors in \(\ell_p^n\)-balls
• scientific article; zbMATH DE number 1713116 (Why is no real title available?)
• scientific article; zbMATH DE number 177384 (Why is no real title available?)
• scientific article; zbMATH DE number 194664 (Why is no real title available?)
• scientific article; zbMATH DE number 6276198 (Why is no real title available?)
• Large deviations for high-dimensional random projections of \(\ell_p^n\)-balls
• Large deviations for random projections of \(\ell^{p}\) balls
• Large deviations techniques and applications.
• Measure theory and fine properties of functions
• Power-law estimates for the central limit theorem for convex sets
• Probability theory. A comprehensive course
• Quenched large deviation principle for words in a letter sequence
• Random measures, theory and applications
• Random projections of regular polytopes
• Random projections of regular simplices
• Regular simplices and Gaussian samples
• Sanov-type large deviations in Schatten classes
• Stability of low-rank matrix recovery and its connections to Banach space geometry
• The central limit problem for random vectors with symmetries
• The curse of dimensionality for numerical integration on general domains
• The dimension of almost spherical sections of convex bodies
• The Gelfand widths of \(\ell_p\)-balls for \(0 < p \leq 1\)

Cited in

(11)

• Thin-shell theory for rotationally invariant random simplices
• Thin-shell concentration for random vectors in Orlicz balls via moderate deviations and Gibbs measures
• Product of random projections, Jacobi ensembles and universality problems arising from free probability
• A central limit theorem for projections of the cube
• Projections of the uniform distribution on the cube: a large deviation perspective
• Limit theorems for the volumes of small codimensional random sections of \(\ell_{p}^{n}\)-balls
• Higher order concentration on Stiefel and Grassmann manifolds
• Large deviation principles induced by the Stiefel manifold, and random multidimensional projections
• Large deviations for random matrices in the orthogonal group and Stiefel manifold with applications to random projections of product distributions
• Limit theorems for Herer expectations of random sets in Banach spaces
• Sharp concentration phenomena in high-dimensional Orlicz balls

This page was built for publication: A new look at random projections of the cube and general product measures

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2040109)