A note on the variance of the square components of a normal multivariate within a Euclidean ball

Abstract: We present arguments in favour of the inequalities

v a r (X_{n}^{2} | X i n B_{v} (h o)) l e 2 l a m b d a_{n} E [X_{n}^{2} | X i n B_{v} (h o)]

, where

X s i m N_{v} (0, L a m b d a)

is a normal vector in

v g e 1

dimensions, with zero mean and covariance matrix

L a m b d a = d i a g (l a m b d a)

, and

B_{v} (h o)

is a centered

v

-dimensional Euclidean ball of square radius

h o

. Such relations lie at the heart of an iterative algorithm, proposed in ref. [1] to perform a reconstruction of

L a m b d a

from the covariance matrix of

X

conditioned to

B_{v} (h o)

. In the regime of strong truncation, i.e. for

h o l e s s s i m l a m b d a_{n}

, the above inequality is easily proved, whereas it becomes harder for

h o g g l a m b d a_{n}

. Here, we expand both sides in a function series controlled by powers of

l a m b d a_{n} / h o

and show that the coefficient functions of the series fulfill the inequality order by order if

h o

is sufficiently large. The intermediate region remains at present an open challenge.

Recommendations

Cites Work

Cited In (4)

This page was built for publication: A note on the variance of the square components of a normal multivariate within a Euclidean ball