Markov Inequalities for Weight Functions of Chebyshev Type
โ Scribed by D.K. Dimitrov
- Publisher
- Elsevier Science
- Year
- 1995
- Tongue
- English
- Weight
- 159 KB
- Volume
- 83
- Category
- Article
- ISSN
- 0021-9045
No coin nor oath required. For personal study only.
โฆ Synopsis
Denote by (\eta_{i}=\cos (i \pi / n), i=0, \ldots, n) the extreme points of the Chebyshev polynomial (T_{n}(x)=\cos (n \operatorname{arc} \cos x)). Let (\pi_{n}) be the set of real algebraic polynomials of degree not exceeding (n), and let (B_{n}) be the unit ball in the space (\pi_{n}) equipped with the discrete norm (|p|{n . x}:=\max {0 \leqslant i \leqslant n}\left|p\left(\eta{i}\right)\right|). We prove that the unique solutions of the extremal problems (\max {p \in B{n}} \int{-1}^{1}\left[p^{\prime k+1)}(x)\right]^{2}\left(1-x^{2}\right)^{k-1 / 2} d x, k=0, \ldots, n-1), and (\max {p \in B{n}} \int_{-1}^{1}\left[p^{1 k+21}(x)\right]^{2}\left(1-x^{2}\right)^{k-1 / 2} d x, k=0, \ldots, n-2), are (p(x)= \pm T_{n}(x)), and we obtain the extremal values in an explicit form. 1995 Academic Press, Inc
๐ SIMILAR VOLUMES
## Abstract The first term of the asymptotics of the best constants in Markovโtype inequalities for higher derivatives of polynomials is determined in the two cases where the underlying norm is the __L__^2^ norm with Laguerre weight or the __L__^2^ norm with Gegenbauer weight. The coefficient in th