Arestov's theorems on Bernstein's inequality
Tam\'as Erd\'elyi

TL;DR
This paper presents a straightforward and partly novel proof of Arestov's extension of Bernstein's inequality in Lp spaces, utilizing an approach related to Mahler's inequality for polynomials.
Contribution
It introduces a simplified proof of Arestov's theorem, extending Boyd's approach from algebraic to trigonometric polynomials.
Findings
Proof of Arestov's extension of Bernstein's inequality in Lp for all p ≥ 0
Extension of Boyd's approach to trigonometric polynomials
Simplification of existing proof techniques
Abstract
We give a simple, elementary, and at least partially new proof of Arestov's famous extension of Bernstein's inequality in to all . Our crucial observation is that Boyd's approach to prove Mahler's inequality for algebraic polynomials can be extended to all trigonometric polynomials .
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Arestov’s theorems on Bernstein’s inequality
Tamás Erdélyi
Department of Mathematics, Texas A&M University, College Station, Texas 77843, College Station, Texas 77843
(April 26, 2019 )
We give a simple, elementary, and at least partially new proof of Arestov’s famous extension of Bernstein’s inequality in to all . Our crucial observation is that Boyd’s approach to prove Mahler’s inequality for algebraic polynomials can be extended to all trigonometric polynomials .
Bernstein’s inequality in for all , Arestov’s theorem
††support: 2010 Mathematics Subject Classifications. 41A17
1. Introduction and Notation
Let be the collection of all trigonometric polynomials of the form
[TABLE]
Let be the collection of all algebraic polynomials of the form
[TABLE]
Let denote the open unit disk, and let denote the unit circle of the complex plane. We define the Mahler measure (geometric mean of on )
[TABLE]
for bounded measurable functions on . It is well known, see [HL-52], for instance, that
[TABLE]
where
[TABLE]
It is also well known that for a function continuous on we have
[TABLE]
It is a simple consequence of the Jensen formula that
[TABLE]
for every polynomial of the form
[TABLE]
Bernstein’s inequality
[TABLE]
plays a crucial role in proving inverse theorems of approximation as well as many other results in approximation theory. See [BE-95], for instance. As far as the history of Bernstein’s inequality is concerned we refer to Nevai’s lovely papers [N-14] and [N-19] and the references in them. We do not repeat the full story here. In 1981 Arestov [A-81] proved that
[TABLE]
for all , extending the result known only for for a long time. Simpler proof of Bernstein’s inequality in for all have been given by Golitschek and Lorentz in [GL-89] which is presented in the book [DL-93] by DeVore and Lorentz. A very elegant and even more simplified proof was published recently in [QZ-19] by Queffélec and Zarouf. A central part of their proof is to prove (1.1) for first. Mahler [M-61] showed (1.1) for but only for polynomials , and he gave a rather involved proof. Mahler’s inequality was also posed as a problem by Vaaler in the Problems section of the American Mathematical Monthly and solved by Boyd [VB-91] using an elementary theorem of Bernstein. We note that Glazyrina [G-05] proved a sharp Markov-type inequality for algebraic polynomials in on finite subintervals of the real line.
In this note we give a simple, elementary, and at least partially new proof of Arestov’s famous extension of Bernstein’s inequality in to all . Our crucial observation is that Boyd’s approach to prove Mahler’s inequality for algebraic polynomials can be extended to all trigonometric polynomials .
Theorem 1.1
We have
[TABLE]
Equivalently
[TABLE]
Theorem 1.2
With the notation we have
[TABLE]
Theorem 1.3
We have
[TABLE]
for every .
2. Lemmas
To prove the Theorem 1.1 we need two lemmas.
Lemma 2.1
Associated with let be defined by . If has each of its zeros in , then has each of its zeros in as well. The same is true if is replaced by the closed unit disk .
Demonstration Proof
We prove the lemma for , the case of the closed unit disk follows from this by a straightforward limiting argument. Suppose , that is, and . Suppose also that and has each of its zeros in , that is
[TABLE]
We have
[TABLE]
Observe that for each , and hence
[TABLE]
Combining (2.1) and (2.2) we obtain
[TABLE]
We conclude that . ∎
Lemma 2.2
Associated with let be defined by and suppose that has each of its zeros in the closed unit disk . If and
[TABLE]
then
[TABLE]
Demonstration Proof
Without loss of generality we may assume that has each of its zeros in , the case when has each of its zeros in the closed unit disk follows from this by a straightforward limiting argument. Let be defined by Let , , and
[TABLE]
It follows from (2.3) that
[TABLE]
and hence and the fact that does not vanish on imply that
[TABLE]
Therefore Rouche’s Theorem implies that the polynomial and has the same number of zeros in , that is, has each of its zeros in . By Lemma 2.1 and (2.4) we can deduce that has each of its zeros in . In particular,
[TABLE]
for all , . We conclude that
[TABLE]
∎
3. Proof of Theorems 1.1, 1.2, and 1.3
Demonstration Proof of Theorem 1.1
Associated with let be defined by . Without loss of generality we may assume that has exactly complex zeros, the case when has less than complex zeros follows from this by a straightforward limiting argument.
Case 1. Suppose has all its zeros in the closed unit disk . It follows from Lemma 2.1 that has all its zeros in the closed unit disk . Let
[TABLE]
Using Jensen’s formula and the multiplicative property of the Mahler measure we can easily deduce that
[TABLE]
Case 2. Suppose that some of the zeros of are outside the closed unit disk . Let be the zeros of outside the closed unit disk and let be the zeros of in the closed unit disk . We have
[TABLE]
We define
[TABLE]
Observe that (2.3) holds, , and defined by has each of its zeros in the closed unit disk . Using Lemma 2.2, the (in)equality of the theorem in Case 1, and Jensen’s formula, we obtain
[TABLE]
∎
Demonstration Proof of Theorem 1.2
We follow the argument given in [QZ-19] to base our proof on Theorem 1.1. It is well-known, and by applying Jensen’s formula it is ieasy to see, that
[TABLE]
and hence
[TABLE]
Let , , and . Applying Theorem 1.1 with replaced by we obtain
[TABLE]
Integrating both sides on with respect to , then using Fubini’s theorem and (3.1), we get the theorem ∎
Demonstration Proof of Theorem 1.3
We follow the argument given in [QZ-19] to base our proof on Theorem 1.2. Observe that
[TABLE]
Indeed, the integration by parts formula gives
[TABLE]
For the sake of brevity we will use the notation . Using (3.2), Fubini’s theorem, Theorem 1.2, and Fubini’s theorem again, we obtain
[TABLE]
∎
4. Acknowledgement
The author thanks Herve Queffélec and Paul Nevai for checking the details of the proof in this paper and for their suggestions to make the paper more readable.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1A-81 V.V. Arestov , On integral inequalities for trigonometric polynomials and their derivatives , Izv. 45 ( 1981 ), 3–22 .
- 2BE-95 P. Borwein and T. Erdélyi , Polynomials and Polynomial Inequalities , Springer , 1995 .
- 3DL-93 R.A. De Vore and G.G. Lorentz , Constructive Approximation , Springer-Verlag , 1993 .
- 4G-05 P. Yu. Glazyrina , The Markov brothers’ inequality in the space L 0 subscript 𝐿 0 L_{0} on an interval. (Russian) , translation in Math. Notes 78 (2005), no. 1-2, 53–58 , Mat. Zametki 78 ( 2005 ), no. 1, 59–65 .
- 5GL-89 M.V. Golitschek and G.G. Lorentz , Bernstein inequalities in L p subscript 𝐿 𝑝 L_{p} , 0 ≤ p ≤ ∞ 0 𝑝 0\leq p\leq\infty , Rocky Mountain J. Math. 19 ( 1989 ), no. 1, 145-156 .
- 6M-61 K. Mahler , On the zeros of the derivative of a polynomial , Proc. Roy. Soc. London Ser. A 264 ( 1961 ), 145–154. .
- 7N-14 P. Nevai and The Anonymous Referee , The Bernstein inequality and the Schur inequality are equivalent. , J. Approx. Theory 182 ( 2014 ), 103–109 .
- 8N-19 P. Nevai , The True Story of n 𝑛 n vs. 2 n 2 𝑛 2n in the Bernstein Inequality , in preparation ( 2019 ).
