Projections of the uniform distribution on the cube: a large deviation perspective

Abstract: Let

T h e t a^{(n)}

be a random vector uniformly distributed on the unit sphere

m a t h b b S^{n - 1}

in

m a t h b b R^{n}

. Consider the projection of the uniform distribution on the cube

[- 1, 1]^{n}

to the line spanned by

T h e t a^{(n)}

. The projected distribution is the random probability measure

m u_{T h e t a^{(n)}}

on

m a t h b b R

given by [ mu_{Theta^{(n)}}(A) := frac 1 {2^n} int_{[-1,1]^n} mathbb 1{langle u, Theta^{(n)} angle in A} du, ] for Borel subets

A

of

m a t h b b R

. It is well known that, with probability

1

, the sequence of random probability measures

m u_{T h e t a^{(n)}}

converges weakly to the centered Gaussian distribution with variance

1 / 3

. We prove a large deviation principle for the sequence

m u_{T h e t a^{(n)}}

on the space of probability measures on

m a t h b b R

with speed

n

. The (good) rate function is explicitly given by

I (u (a l p h a)) := - f r a c 12 l o g (1 - | a l p h a |_{2}^{2})

whenever

u (a l p h a)

is the law of a random variable of the form �egin{align*} sqrt{1 - |alpha|_2^2 } frac{Z}{sqrt 3} + sum_{ k = 1}^infty alpha_k U_k, end{align*} where

Z

is standard Gaussian independent of

U_{1}, U_{2}, l d o t s

which are i.i.d.

e x t U n i f [- 1, 1]

, and

a l p h a_{1} g e q a l p h a_{2} g e q l d o t s

is a non-increasing sequence of non-negative reals with

| a l p h a |_{2} < 1

. We obtain a similar result for random projections of the uniform distribution on the discrete cube

{- 1, + 1}^{n}

.

Recommendations

Cites work

Cited in

(4)

This page was built for publication: Projections of the uniform distribution on the cube: a large deviation perspective