An asymptotic thin shell condition and large deviations for random multidimensional projections

DOI10.1016/J.AAM.2021.102306MaRDI QIDQ2070086zbMATH OpenOpenAlexFDO

Authors Steven Soojin Kim, Yin-Ting Liao, Kavita Ramanan

Publication date 21 January 2022

Published in Advances in Applied Mathematics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1912.13447

zbMATH Keywords

\(\ell_p^n\) ballsasymptotic thin shell conditioncentral limit theorem for convex setsGibbs measuresKLS conjecturelarge deviationsOrlicz ballsrandom projectionsrate functionStiefel manifoldthin shell condition

Mathematics Subject Classification ID

Asymptotic theory of Banach spaces (46B06) Asymptotic theory of convex bodies (52A23) Large deviations (60F10)

Abstract: Consider the projection of an $n$-dimensional random vector onto a random $k_n$-dimensional basis, $k_{n}leqn$, drawn uniformly from the Haar measure on the Stiefel manifold of orthonormal $k_n$-frames in $mathbb{R}^n$, in three different asymptotic regimes as $nightarrowinfty$: "constant" ($k_n=k$), "sublinear" ($k_{n}ightarrowinfty$ but $k_n/nightarrow0$) and "linear" $k_n/nightarrowlambda$ with $0<lambdale1$). When the sequence of random vectors satisfies a certain "asymptotic thin shell condition", we establish annealed large deviation principles (LDPs) for the corresponding sequence of random projections in the constant regime, and for the sequence of empirical measures of the coordinates of the random projections in the sublinear and linear regimes. We also establish LDPs for certain scaled $el{l}_q$ norms of the random projections in these different regimes. Moreover, we verify our assumptions for various sequences of random vectors of interest, including those distributed according to Gibbs measures with superquadratic interaction potential, or the uniform measure on suitably scaled $el{l}_p^n$ balls, for $pin[1,infty)$, and generalized Orlicz balls defined via a superquadratic function. Our results complement the central limit theorem for convex sets and related results which are known to hold under a "thin shell" condition. These results also substantially extend existing large deviation results for random projections, which are first, restricted to the setting of measures on $el{l}_p^n$ balls, and secondly, limited to univariate LDPs (i.e., in $mathbbR$) involving either the norm of a $k_n$-dimensional projection or the projection of $X^{(n)}$ onto a random one-dimensional subspace. Random projections of high-dimensional random vectors are of interest in a range of fields including asymptotic convex geometry and high-dimensional statistics.

Recommendations

• Large deviations for high-dimensional random projections of \(\ell_p^n\)-balls
• Large deviations for random projections of \(\ell^{p}\) balls
• Thin-shell concentration for random vectors in Orlicz balls via moderate deviations and Gibbs measures
• Gaussian fluctuations for high-dimensional random projections of \(\ell_p^n\)-balls
• High-dimensional limit theorems for random vectors in \(\ell_p^n\)-balls. II

Cites work

• A central limit theorem for convex sets
• A conditional limit theorem for high-dimensional ^p-spheres
• A note on a method for generating points uniformly on n -dimensional spheres
• A probabilistic approach to the geometry of the \(\ell^n_p\)-ball
• A Projection Pursuit Algorithm for Exploratory Data Analysis
• A simplified proof of CLT for convex bodies
• Aging of spherical spin glasses
• Approaching the Kannan-Lovász-Simonovits and variance conjectures
• Approximate independence of distributions on spheres and their stability properties
• Asymptotics of cross sections for convex bodies
• Asymptotics of graphical projection pursuit
• Concentration of mass on convex bodies
• Cramér's theorem is atypical
• Descartes' Rule of Signs Revisited
• Extensions of Lipschitz mappings into a Hilbert space
• Generalized Dirichlet distributions on the ball and moments
• Handbook of stochastic analysis and applications
• High-dimensional limit theorems for random vectors in \(\ell_p^n\)-balls
• High-dimensional limit theorems for random vectors in \(\ell_p^n\)-balls. II
• scientific article; zbMATH DE number 3647821 (Why is no real title available?)
• scientific article; zbMATH DE number 1775450 (Why is no real title available?)
• Integral Limit Theorems Taking Large Deviations into Account when Cramér’s Condition Does Not Hold. I
• Interpolating thin-shell and sharp large-deviation estimates for isotropic log-concave measures
• Isoperimetric problems for convex bodies and a localization lemma
• Large deviations for configurations generated by Gibbs distributions with energy functionals consisting of singular interaction and weakly confining potentials
• Large deviations for empirical measures of mean-field Gibbs measures
• Large deviations for exchangeable random vectors
• Large deviations for high-dimensional random projections of \(\ell_p^n\)-balls
• Large deviations for random projections of \(\ell^{p}\) balls
• Large deviations, moderate deviations, and the KLS conjecture
• Latent semantic indexing: A probabilistic analysis
• On Deviations of the Sample Mean
• On the Volume of the Intersection of Two L n p Balls
• Orlicz spaces and modular spaces
• Power-law estimates for the central limit theorem for convex sets
• Projections of probability distributions: a measure-theoretic Dvoretzky theorem
• Sanov property, generalized I-projection and a conditional limit theorem
• Sanov's theorem in the Wasserstein distance: a necessary and sufficient condition
• Sudakov's typical marginals, random linear functionals and a conditional central limit theorem
• The central limit problem for convex bodies
• The KLS isoperimetric conjecture for generalized Orlicz balls
• The maximum entropy principle and volumetric properties of Orlicz balls
• The negative association property for the absolute values of random variables equidistributed on a generalized Orlicz ball
• Thin-shell concentration for random vectors in Orlicz balls via moderate deviations and Gibbs measures

Cited in

(14)

• Large deviations for high-dimensional random projections of \(\ell_p^n\)-balls
• Thin-shell theory for rotationally invariant random simplices
• Thin-shell concentration for random vectors in Orlicz balls via moderate deviations and Gibbs measures
• Large deviations for uniform projections of $p$-radial distributions on $\ell_p^n$-balls
• Sanov-type large deviations and conditional limit theorems for high-dimensional Orlicz balls
• A Maxwell principle for generalized Orlicz balls
• Large deviation principles induced by the Stiefel manifold, and random multidimensional projections
• Geometric sharp large deviations for random projections of _p^n spheres and balls
• Large deviations for random matrices in the orthogonal group and Stiefel manifold with applications to random projections of product distributions
• A probabilistic approach to Lorentz balls \(\ell_{q, 1}^n\)
• Sharp concentration phenomena in high-dimensional Orlicz balls
• Random projections beyond zero overlap
• Selective inference using randomized group Lasso estimators for general models
• Quenched large deviation principles for random projections of _p^n balls

This page was built for publication: An asymptotic thin shell condition and large deviations for random multidimensional projections

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2070086)