On the randomized complexity of Banach space valued integration

Abstract

We study the complexity of Banach space valued integration in the randomized setting. We are concerned with $r$-times continuously differentiable functions on the $d$-dimensional unit cube $Q$, with values in a Banach space $X$, and investigate the relation of the optimal convergence rate to the geometry of $X$. It turns out that the $n$-th minimal errors are bounded by $c{n}^{-r/d-1+1/p}$ if and only if $X$ is of equal norm type $p$.

12 pages