On sequences of polynomials arising from graph invariants

TL;DR

This paper investigates sequences of graph polynomials satisfying linear recurrence relations, providing new characterizations of graph families and exploring their algebraic and invariant properties.

Contribution

It establishes a general theorem for graph families characterized by polynomial recurrences and discusses applications beyond classical Hermite and Laguerre polynomials.

Findings

Characterization of graph families via polynomial recurrence relations

Extension to recurrence relations with constant coefficients

Insights into the algebraic and invariant properties of graph polynomials

Abstract

Graph polynomials are deemed useful if they give rise to algebraic characterizations of various graph properties, and their evaluations encode many other graph invariants. Algebraic: The complete graphs $K_n$ and the complete bipartite graphs $K_{n,n}$ can be characterized as those graphs whose matching polynomials satisfy a certain recurrence relations and are related to the Hermite and Laguerre polynomials. An encoded graph invariant: The absolute value of the chromatic polynomial $\chi(G,X)$ of a graph $G$ evaluated at $-1$ counts the number of acyclic orientations of $G$. In this paper we prove a general theorem on graph families which are characterized by families of polynomials satisfying linear recurrence relations. This gives infinitely many instances similar to the characterization of $K_{n,n}$. We also show where to use, instead of the Hermite and Laguerre polynomials, linear recurrence relations where the coefficients do not depend on $n$. Finally, we discuss the distinctive power of graph polynomials in specific form.