Advanced Search

An isomorphism between some anisotropic Besov and sequence spaces is established, and the continuity of a Stieltjes-type integral operator, acting on some of these spaces, is proved.

The asymptotic behaviour of ε-entropy of classes of Lipschitz functions in $L^p{(}^d)$ is obtained. Moreover, the asymptotics of ε-entropy of classes of Lipschitz functions in $L^p\left({ℝ}^d\right)$ whose tail function decreases as $O\left({\lambda }^{-\gamma }\right)$ is obtained. In case p = 1 the relation between the ε-entropy of a given class of probability densities on ${ℝ}^d$ and the minimax risk for that class is discussed.

To each set of knots $t_i=i/2n$ for i = 0,...,2ν and $t_i=(i-\nu )/n$ for i = 2ν + 1,..., n + ν, with 1 ≤ ν ≤ n, there corresponds the space ${}_{\nu ,n}$ of all piecewise linear and continuous functions on I = [0,1] with knots $t_i$ and the orthogonal projection $P_{\nu ,n}$ of L²(I) onto ${}_{\nu ,n}$. The main result is $li{m}_{(n-\nu )\wedge \nu \to \infty }\left|\right|P_{\nu ,n}\left|\right|₁=su{p}_{\nu ,n:1\le \nu \le n}\left|\right|P_{\nu ,n}\left|\right|₁=2+(2-\surd 3)²$. This shows that the Lebesgue constant for the Franklin orthogonal system is 2 + (2-√3)².

We study nonlinear m-term approximation in a Banach space with regard to a basis. It is known that in the case of a greedy basis (like the Haar basis in $L_p\left([0,1]\right)$, 1 < p < ∞) a greedy type algorithm realizes nearly best m-term approximation for any individual function. In this paper we generalize this result in two directions. First, instead of a greedy algorithm we consider a weak greedy algorithm. Second, we study in detail unconditional nongreedy bases (like the multivariate Haar basis ${}^d=\times ...\times$ in $L_p\left({[0,1]}^d\right)$,...