Closeness of convolutions of probability measures

TL;DR

This paper establishes new explicit bounds for the total variation distance between convolution products of probability measures, improving approximation accuracy, especially for symmetric distributions, with applications to multinomial distribution approximation.

Contribution

It introduces novel bounds with magic factors for convolutions of probability measures and improves existing bounds in multinomial approximation, dropping dimension factors.

Findings

New bounds with magic factors for convolutions

Improved multinomial approximation bounds

Better accuracy for symmetric distributions with finite support

Abstract

We derive new explicit bounds for the total variation distance between two convolution products of $n$ probability distributions, one of which having identical convolution factors. Approximations by finite signed measures of arbitrary order are considered as well. We are interested in bounds with magic factors, i.e. roughly speaking $n$ also appears in the denominator. Special emphasis is given to the approximation by the $n$-fold convolution of the arithmetic mean of the distributions under consideration. As an application, we consider the multinomial approximation of the generalized multinomial distribution. It turns out that here the order of some bounds given in Roos (2001) and Loh (1992) can significantly be improved. In particular, it follows that a dimension factor can be dropped. Moreover, better accuracy is achieved in the context of symmetric distributions with finite support. In the course of proof, we use a basic Banach algebra technique for measures on a measurable Abelian group. Though this method was already used by Le Cam (1960), our central arguments seem to be new. We also derive new smoothness bounds for convolutions of probability distributions, which might be of independent interest.