Abstract:It is well known that fluctuations of marginals of high-dimensional random vectors that satisfy a certain concentration estimate called the thin shell condition are approximately Gaussian. In this article we identify a general condition on a sequence of high-dimensional random vectors under which one can identify the exponential decay rate of large deviation probabilities of the corresponding sequence of marginals. More precisely, consider the projection of an n-dimensional random vector onto a random k n -dimensional basis, k n ≤ n , drawn uniformly from the Haar measure on the Stiefel manifold of orthonormal k n -frames in R n , in three different asymptotic regimes as n → ∞ : “constant” ( k n = k ), “sublinear” ( k n → ∞ but k n / n → 0 ) and “linear” ( k n / n → λ with 0 < λ ≤ 1 ). When the sequence of random vectors satisfies a certain “asymptotic thin shell condition”, we establish large deviation principles 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 large deviation principles for scaled ℓ q norms of the random projections in all three regimes. Moreover, we show that the asymptotic thin shell condition holds for various sequences of random vectors of interest, including the uniform measure on suitably scaled ℓ p n balls, for p ∈ [ 1 , ∞ ) , and generalized Orlicz balls defined via a superquadratic function, as well as a class of Gibbs measures with superquadratic interaction potential. Along the way, we obtain logarithmic asymptotics of volumes of high-dimensional Orlicz balls, which may be of independent interest. We also show that the decay rate of large deviation probabilities of Euclidean norms of multi-dimensional projections of ℓ p n balls, when p ∈ [ 1 , 2 ) , exhibits an unexpected phase transition in the sublinear regime, thus disproving an earlier conjecture due to Alonso-Gutiérrez et al. Random projections of high-dimensional random vectors are of interest in a range of fields including asymptotic convex geometry and high-dimensional statistics.

Thin-shell concentration for random vectors in Orlicz balls via moderate deviations and Gibbs measures

Sharp concentration phenomena in high-dimensional Orlicz balls

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

Volume properties of high-dimensional Orlicz balls

Thin-shell theory for rotationally invariant random simplices

The large and moderate deviations approach in geometric functional analysis

A probabilistic approach to Lorentz balls

Borell's inequality and mean width of random polytopes via discrete inequalities

Quenched large deviation principles for random projections of $\ell_p^n$ balls

Some sharp inequalities of Mizohata--Takeuchi-type

Limit Theorems for the Volume of Random Projections and Sections of $\ell_p^N$-balls

Concentration of Randomized Functions of Uniformly Bounded Variation

The smallest singular value for rectangular random matrices with Lévy entries

Concentration Inequalities for General Functions of Heavy-Tailed Random Variables

Brownian loops on non-smooth surfaces and the Polyakov-Alvarez formula

Exit laws from large balls of (an)isotropic random walks in random environment

Off-Lattice Random Walks with Excluded Volume: A New Method of Generation, Proof of Ergodicity and Numerical Results

Optimal constants in concentration inequalities on the sphere and in the Gauss space

Asymptotic Normality and Concentration Inequalities of Statistics of Core Partitions with Bounded Perimeters

Strange shadows of $\ell_p$-balls

Strichartz estimates for orthonormal functions and probabilistic convergence of density functions of compact operators on manifolds