Simultaneous orthogonal polynomials related to Poisson distribution

From MaRDI portal





Given a positive \(a\), let \(\mu(a;x)\) denote the discrete measure having the masses \(a^k/k!\) at integer nonnegative points \(k\). Let us fix \(a_1,\dots,a_r\) such that \(0<a_1<a_2<\dots<a_r\) and define \(\mu_j(x)=\mu(a_j;x)\), \(j=1,\dots,r\). Suppose that \(Q_n(x)\) is a (nonzero) polynomial of degree \(\deg Q_n \leq rn\), satisfying the orthogonality relations: \[ \int Q_n(x)\,x^m\,d\mu_j(x)=0, \quad m=0,\dots,n-1, \quad j=1,\dots,r. \] Further, let \(\{\gamma_n\}\) be some nondecreasing sequence of positive numbers called scaling coefficients. Scaled simultaneous Charlier polynomials are defined to be the polynomials \[ Q^*_n(x) =C_nQ_n(\gamma_nx), \quad n\in \mathbb{Z}_+, \] where the normalizing factor \(C_n\) is taken so that the leading coefficient of \(Q^*_n\) is equal to one. By the saddle-point method it is shown that the limit measure of distribution of zeros of scaled simultaneous Charlier polynomials is the classic Lebesque measure on the segment \([0, r]\). A similar result is obtained for the so-called scaled generalized Charlier polynomials.











This page was built for publication: Simultaneous orthogonal polynomials related to Poisson distribution

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q493878)