Widths of weighted Sobolev classes on the ball (Q999270): Difference between revisions

\textit{F. Dai} [J. Math. Anal. Appl. 315, No. 2, 711--724 (2006; Zbl 1094.41015)] has recently studied the Kolmogorov \(n\)-widths \(d_n\) and the linear \(n\)-widths \(\delta_n\) of the weighted Besov classes \(B^{\alpha}_s(L_{p,\mu})\) on [-1,1] in \(L_{q,\mu}\), where \(L_{q,\mu}\;(1\leq p,q\leq \infty)\) denotes the weighted \(L_q\) space of functions on [-1,1] with respect to weight \((1-t^2)^{\mu-\frac12}\;(\mu>0)\) and has given optimal asymptotic orders of \(d_n(B^{\alpha}_s(L_{p,\mu}),L_{q,\mu})\) and \(\delta_n(B^{\alpha}_s(L_{p,\mu}),L_{q,\mu})\) as \(n\to\infty\). The main purpose of this paper is to establish multivariate analogues of these results on the unit ball in \(\mathbb R^d\) and to obtain optimal asymptotic estimates about the Kolmogorov and linear widths of weighted Sobolev classes \(BW^r_{p,\mu}\) in \(L_{q,\mu}\) for all \(1\leq p,q\leq\infty\) and \(\mu\geq 0\). The details are too technical to be reproduced here.