[Paper Review] C*-algebraic Schoenberg Conjecture

K. Mahesh Krishna|arXiv (Cornell University)|Jun 14, 2022

Advanced Topics in Algebra4 citations

TL;DR

This paper formulates a C*-algebraic generalization of Schoenberg's conjecture on the spectral localization of polynomial derivatives, extending classical complex polynomial inequalities to noncommutative C*-algebraic settings. It proves the conjecture holds for degree 2 polynomials, establishing operator norm inequalities involving sums of $ b_k b_k^* $ and $ b_k^* b_k $ in terms of the original coefficients $ a_j $ and their adjoints.

ABSTRACT

Based on Schoenberg conjecture extit{[Amer. Math. Monthly., 1986]}/Malamud-Pereira theorem extit{[J. Math. Anal. Appl, 2003]}, extit{[Trans. Amer. Math. Soc., 2005]} we formulate the following conjecture which we call C*-algebraic Schoenberg Conjecture.\\ extbf{ C*-algebraic Schoenberg Conjecture : Let $\mathcal{A}$ be a C*-algebra. Let $d\in \mathbb{N}\setminus\{1\}$, $P(z)= (z-a_1)(z-a_2)\cdots (z-a_d)$ be a polynomial over $\mathcal{A}$ with $a_1, a_2, \dots, a_d \in \mathcal{A} $. If $P'$ can be written as $P'(z)= d(z-b_1)(z-b_2)\cdots (z-b_{d-1})$ on $\mathcal{A}$ with $b_1, b_2, \dots, b_{d-1} \in \mathcal{A} $, then \begin{align*} \sum_{k=1}^{d-1}b_kb_k^*\leq \frac{1}{d^2}\left[\sum_{j=1}^{d}a_j ight]\left[\sum_{j=1}^{d}a_j ight]^*+ \frac{d-2}{d}\sum_{j=1}^{d}a_ja_j^* \end{align*} and \begin{align*} \sum_{k=1}^{d-1}b_k^*b_k\leq \frac{1}{d^2}\left[\sum_{j=1}^{d}a_j ight]^*\left[\sum_{j=1}^{d}a_j ight]+ \frac{d-2}{d}\sum_{j=1}^{d}a_j^*a_j. \end{align*}} We show that C*-algebraic Schoenberg conjecture holds for degree 2 C*-algebraic polynomials over C*-algebras.

Motivation & Objective

• To extend Schoenberg's conjecture on polynomial roots and derivatives to the noncommutative setting of C*-algebras.
• To formulate operator-theoretic analogues of classical inequalities involving sums of squared moduli of critical points.
• To investigate whether the classical inequalities for complex polynomials generalize to C*-algebra-valued polynomials under appropriate algebraic and spectral conditions.
• To establish a C*-algebraic version of the de Bruin-Sharma and Kushel-Tyaglov inequalities as conjectures.
• To verify the validity of the C*-algebraic Schoenberg conjecture in the simplest nontrivial case: degree 2 polynomials.

Proposed method

• Define a C*-algebraic polynomial $ P(z) = (z - a_1)igcdots(z - a_d) $ with $ a_j otin ext{center}( ext{A}) $, allowing noncommutative coefficients.
• Define the derivative $ P'(z) $ via the standard product rule in the algebraic sense, with $ P'(z) = ext{sum over missing factors} $, and assume it factors as $ d(z - b_1)igcdots(z - b_{d-1}) $.
• Derive operator inequalities comparing $ ext{Tr}( ext{sum } b_k b_k^*) $ and $ ext{Tr}( ext{sum } b_k^* b_k) $ to expressions involving $ ext{sum } a_j $, $ ext{sum } a_j a_j^* $, and $ ext{sum } a_j^* a_j $.
• Use the degree 2 case to explicitly compute $ b_1 = (a + b)/2 $, then verify the conjectured inequality holds with equality.
• Formulate C*-algebraic analogues of the de Bruin-Sharma and Kushel-Tyaglov theorems as conjectures, involving higher moments of $ b_k $ and $ a_j $.
• Apply standard C*-algebra identities, including $ (a+b)(a^*+b^*) = aa^* + ab^* + ba^* + bb^* $, to expand and compare both sides of the inequalities.

Experimental results

Research questions

• RQ1 Can Schoenberg’s classical inequality on the sum of squared moduli of critical points be generalized to C*-algebra-valued polynomials?
• RQ2 Does the C*-algebraic version of the inequality hold with equality in the degree 2 case, and if so, under what conditions?
• RQ3 What are the appropriate noncommutative analogues of the de Bruin-Sharma and Kushel-Tyaglov inequalities in the C*-algebra setting?
• RQ4 How do the spectral properties of the critical points $ b_k $ relate to the coefficients $ a_j $ in a noncommutative algebra?
• RQ5 Under what algebraic and analytic conditions can the derivative of a C*-polynomial be factored into linear terms over the algebra?

Key findings

• The C*-algebraic Schoenberg conjecture holds with equality for all degree 2 polynomials over any C*-algebra.
• For degree 2, the critical point is $ b_1 = (a + b)/2 $, and the inequality $ b_1 b_1^* riangleq rac{(a+b)(a^*+b^*)}{4} $ matches the right-hand side of the conjectured bound.
• The equality case arises because the correction terms vanish when $ d = 2 $, reducing the general inequality to a tautological expansion.
• The conjectured C*-algebraic version of the de Bruin-Sharma inequality holds trivially for degree 2, as $ d-4 = -2 $, and the structure of the terms simplifies.
• The C*-algebraic Kushel-Tyaglov conjecture reduces to a nontrivial but verifiable inequality in degree 2, where all higher-order terms are explicitly computable.
• The paper establishes a framework for extending classical complex polynomial inequalities to noncommutative operator algebras, with a complete proof in the minimal nontrivial case.