Optimal algorithms for numerical integration: recent results and open problems (Q7012124)

Let \( D_d\subset\mathbb R^d \) be open set with Lebesgue measure 1. For example one may take the unit cube \(D_d:=[0,1]^d\). The subject of this paper is studying algorithms and tractibility for the the approximation of the integral \(S_d(f)=\int_{D_d}f(x)dx\) for integrable functions \(f:D_d\to \mathbb R\). Given fixed \( \varepsilon>0\) the underlying question concerns the minimal number of function values that are required to achieve an error of \(\varepsilon \). The author presents a detailed and comprehensive survey on optimal algorithms for numerical integration in the following senses. Lower bounds, universality, general domains and IID information. There is an excellent set of open problems presented with some well written concluding remarks. As a world authority on this subject, the author has produced an up to date excellent and well written paper with a wealth of good references.\N\NFor the entire collection see [Zbl 1552.65006].