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.

Normal Projection: Deterministic and Probabilistic Algorithms

Which exceptional low-dimensional projections of a Gaussian point cloud can be found in polynomial time?

Random projections of random manifolds

Spherical random projection

Automatic Least-Squares Projection of Points Onto Point Clouds with Applications in Reverse Engineering

On Degenerate Doubly Nonnegative Projection Problems

Invariance principle of random projection for the norm

Probability Tools for Sequential Random Projection

Projection Test for Mean Vector in High Dimensions

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

Improved Algorithm and Bounds for Successive Projection

Proving Inequalities and Solving Global Optimization Problems Via Simplified CAD Projection

Random projections beyond zero overlap

Applications Studies of Projection Pursuit in the Chance Constrained Programming

Uncertain random geometric programming problems

Directional Differentiability of the Metric Projection in Bochner Spaces

Normal Estimation on Manifolds by Gradient Learning.

Random Projections of Sparse Adjacency Matrices

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

Subspace Principal Angle Preserving Property Of Gaussian Random Projection

On the mean projection theorem for determinantal point processes