The value at the mode in multivariate t distributions: a curiosity or not?

Publication date: 6 November 2012

Abstract: It is a well-known fact that multivariate Student

t

distributions converge to multivariate Gaussian distributions as the number of degrees of freedom

u

tends to infinity, irrespective of the dimension

k g e q 1

. In particular, the Student's value at the mode (that is, the normalizing constant obtained by evaluating the density at the center)

c_{u, k} = f r a c G a m m a (f r a c u + k 2) (p i u)^{k / 2} G a m m a (f r a c u 2)

converges towards the Gaussian value at the mode

c_{k} = f r a c 1 (2 p i)^{k / 2}

. In this note, we prove a curious fact:

c_{u, k}

tends monotonically to

c_{k}

for each

k

, but the monotonicity changes from increasing in dimension

k = 1

to decreasing in dimensions

k g e q 3

whilst being constant in dimension

k = 2

. A brief discussion raises the question whether this emph{a priori} curious finding is a curiosity, emph{in fine}.

This page was built for publication: The value at the mode in multivariate $t$ distributions: a curiosity or not?