Quadrature for self-affine distributions on \(\mathbb R^d\) (Q895698): Difference between revisions

In this interesting paper, the authors present a systematic treatment of quadrature problems with respect to fractal probability measures for function classes of fractional smoothness of Hölder-Lipschitz type. The task is to compute an approximation to the integral \[ I(f) = \int f\,dP \] for a self-similar probability distribution \(P\) on \({\mathbb R}^d\) \((d\in {\mathbb N})\) and a function \(f:\,{\mathbb R}^d \to {\mathbb R}\) by means of a deterministic, resp. randomized algorithm that is based on finitely many evaluations of \(f\). The authors assume that \(P\) is self-similar with respect to a finite number of affine contractions and that \(f\) satisfies a smoothness condition of Hölder-Lipschitz type. Deterministic and randomized composite quadrature rules are presented. The corresponding errors are studied. Further, the authors derive lower bounds for worst-case errors of arbitrary integration schemes that show optimality of the proposed algorithms in many cases.