1: \begin{abstract}

2: We develop a convergence analysis of a multi-level algorithm combining

3: higher order quasi-Monte Carlo (QMC) quadratures with general

4: Petrov-Galerkin discretizations of countably affine parametric operator

5: equations of elliptic and parabolic type, extending both the multi-level

6: first order analysis in [\emph{F.Y.~Kuo, Ch.~Schwab, and I.H.~Sloan,

7: Multi-level quasi-Monte Carlo finite element methods for a class of

8: elliptic partial differential equations with random coefficient}

9: ({Found.\ Comp.\ Math., 2015})] and the single level higher order

10: analysis in [\emph{J.~Dick, F.Y.~Kuo, Q.T.~Le~Gia, D.~Nuyens, and

11: Ch.~Schwab, Higher order QMC Galerkin discretization for parametric

12: operator equations} ({SIAM J.\ Numer.\ Anal., 2014})]. We cover, in

13: particular, both definite as well as indefinite, strongly elliptic systems

14: of partial differential equations (PDEs) in non-smooth domains, and

15: discuss in detail the impact of higher order derivatives of {\KL}

16: eigenfunctions in the parametrization of random PDE inputs on the

17: convergence results. Based on our \emph{a-priori} error bounds, concrete

18: choices of algorithm parameters are proposed in order to achieve a

19: prescribed accuracy under minimal computational work. Problem classes and

20: sufficient conditions on data are identified where multi-level higher

21: order QMC Petrov-Galerkin algorithms outperform the corresponding single

22: level versions of these algorithms. Numerical experiments confirm the theoretical results.

23: \end{abstract}

24: