On measures induced by extremal signatures in best real polynomial approximation
Let \(f\) be a real-valued continuous function on \([-1,1]\). By the Chebyshev equioscillation theorem there exists a set \(A_n \subset [-1,1]\) of \(n+2\) alternation points for the error function \(f-p^*_n\), where \(p^*_n\) denotes the algebraic polynomial of degree \(\leq n\) of best uniform approximation to \(f\). In this paper, the author considers measures \(\mu_n\) on \(A_n\) induced by extremal signatures for \(f-p^*_n\) that turn out to be more sensitive to the distribution of the alternation points than the (usually considered) point-counting measures \(\nu_n\) on \(A_n\). While, as is well-known, for every \(f\) as above at least a subsequence of \((\nu_n)\) converges, in the weak-star topology, to the equilibrium distribution \(\mu_{[-1,1]}\) of \([-1,1]\), the limit distributions of the sequences \((\mu_n)\) reflect analytic properties of the function \(f\). Roughly speaking, the measures \(\mu_n\) tend to concentrate where \(f\) behaves irregularily.
- On the distribution of alternation points in uniform polynomial approximation of entire functions
- Distribution of Alternation Points in Uniform Polynomial Approximation
- Distribution of alternation points in best rational approximations
- Sequences with equi-distributed extreme points in uniform polynomial approximation
This page was built for publication: On measures induced by extremal signatures in best real polynomial approximation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1381497)
