1: \begin{abstract}

2: In a recent paper Keister proposed two quadrature rules as alternatives

3: to Monte Carlo for certain multidimensional integrals and reported his test

4: results. In earlier work we had shown that the quasi-Monte Carlo method

5: with generalized Faure points is very effective for a variety of high

6: dimensional integrals occuring in mathematical finance. In this paper we

7: report test results of this method on Keister's examples of dimension

8: $9$ and $25$, and also for examples of dimension $60$, $80$ and $100$.

9: 

10: For the $25$ dimensional integral we achieved accuracy of $10^{-2}$ with

11: less than $500$ points while the two methods tested by Keister used

12: more than $220,000$ points. In all of our tests, for $n$ sample points 

13: we obtained an empirical

14: convergence rate proportional to $n^{-1}$ rather than the $n^{-1/2}$ of

15: Monte Carlo.

16: 

17: \end{abstract}

18: