Reproducing kernel orthogonal polynomials on the multinomial distribution

TL;DR

This paper derives reproducing kernel orthogonal polynomials on the multinomial distribution, explores their probabilistic properties, and applies them to goodness-of-fit testing and Markov chain analysis.

Contribution

It introduces a new class of reproducing kernel orthogonal polynomials on the multinomial distribution and provides a duplication formula linking them to Krawtchouk polynomials.

Findings

Derived reproducing kernel orthogonal polynomials for multinomial distribution

Developed a multinomial goodness-of-fit test using orthogonal components

Analyzed cutoff times in Markov chains with multinomial stationary distribution

Abstract

Diaconis and Griffiths (2014) study the multivariate Krawtchouk polynomials orthogonal on the multinomial distribution. In this paper we derive the reproducing kernel orthogonal polynomials Q_n(x,y};N,p) on the multinomial distribution which are sums of products of orthonormal polynomials in x and y of fixed total degree n=0,1,.., N. sum_{n=0}^N rho^nQ_n(x,y);N,p) arises naturally from a probabilistic argument. An application to a multinomial goodness of fit test is developed, where the chi-squared test statistic is decomposed into orthogonal components which test the order of fit. A new duplication formula for the reproducing kernel polynomials in terms of the 1-dimensional Krawtchouk polynomials is derived. The duplication formula allows a Lancaster characterization of all reversible Markov chains with a multinomial stationary distribution whose eigenvectors are multivariate Krawtchouk polynomials and where eigenvalues are repeated within the same total degree. The \chi^2 cutoff time, and total variation cutoff time is investigated in such chains. Emphasis throughout the paper is on a probabilistic understanding of the polynomials and their applications, particularly to Markov chains.