Near best tree approximation (Q1599367)

Tree approximation is a new form of nonlinear wavelet approximation. The distinction between tree approximation and the familiar \(n\)-term wavelet approximation is that the wavelets appearing in the approximant are required to align themselves in a certain connected tree structure. This allows for efficient encoding of the positions of the wavelets used in the approximation. Previous work of \textit{A. Cohen}, \textit{W. Dahmen}, \textit{I. Daubechies} and \textit{R. DeVore} [Appl. Comput. Harmon. Anal. 11, No. 2, 192-226 (2001; Zbl 0992.65151)] has given upper bounds for the error of tree approximation for certain Besov spaces. In this interesting paper, the authors study the tree approximation of individual functions. The approximation error is measured in \(L_2(\mathbb{R}^d)\) and in Besov spaces. The new characterization of functions with a prescribed approximation order is given in terms of a certain maximal function applied to the wavelet coefficients.