Multiplier sequences for simple sets of polynomials (Q2392062): Difference between revisions

One of the fundamental problems in the theory of distribution of zeros of entire functions is to characterize the linear operators, \(T\), which preserve the locus of zeros. Let \(Q=\{q_k (x)\}^\infty_{k=0}\) be a simple set of polynomials (i.e., \(\deg q_k(x) = k\), \(k=0,1,2,\dots)\) and let \(\{\gamma_k\}^\infty_{k =0}\) be a sequence of real numbers. Let \(T : \mathbb R[x]\to \mathbb R[x]\) denote a linear operator associated with the sequence \(\{\gamma_k \}^\infty_{k=0}\), defined by \(T[q_k(x)] :=\gamma_k q_k(x)\) for \(k=0,1,2,\dots\)\,. The operator \(T\) is termed a \(Q\)-multiplier sequence if the polynomial \(T[p(x)]\) has only real zeros whenever the polynomial \(p(x)\) has only real zeros. In a seminal work, \textit{G.~Pólya} and \textit{I.~Schur} [J. für Math. 144, 89--113 (1914; JFM 45.0176.01)] obtained both algebraic and transcendental characterizations of the multiplier sequences when \(Q\) is the standard basis. In the paper under review, the authors solve a decade-old open problem; that is, they characterize the \(Q\)-multiplier sequences for an arbitrary simple set of polynomials. The proof of the main result (Theorem~2) of this interesting paper, while elementary, is truly ingenious.