Monte Carlo and Quasi-Monte Carlo Methods 2010 pp 557-572 | Cite as

# Accelerating the Convergence of Lattice Methods by Importance Sampling-Based Transformations

## Abstract

Importance sampling is a powerful technique for improving the stochastic solution of quadrature problems as well as problems associated with the solution of integral equations, and a generalization of importance sampling, called weighted importance sampling, provides even more potential for error reduction. Additionally, lattice methods are particularly effective for integrating sufficiently smooth periodic functions. We will discuss the advantage of combining these ideas to transform non-periodic to periodic integrands over the unit hypercube to improve the convergence rates of lattice-based quadrature formulas. We provide a pair of examples that show that with the proper choice of importance transformation, the order in the rate of convergence of a quadrature formula can be increased significantly. This technique becomes even more effective when implemented using a family of multidimensional dyadic sequences generally called extensible lattices. Based on an extension of an idea of Soboĺ [17] extensible lattices are both infinite and at the same time return to lattice-based methods with the appropriate choice of sample size. The effectiveness of these sequences, both theoretically and with numerical results, is discussed. Also, there is an interesting parallel with low discrepancy sequences generated by the fractional parts of integer multiples of irrationals which may point the way to a useful construction method for extensible lattices.

## Keywords

Monte Carlo Generate Vector Importance Sampling Lattice Rule Importance Function## Notes

### Acknowledgements

The first two authors wish to dedicate this paper to their friend and mentor, Dr. Jerome Spanier on the occasion of his 80th birthday. The last author gratefully acknowledges partial support from the Laser Microbeam and Medical Program (LAMMP), an NIH Biomedical Technology Resource Center (P41-RR01192). The authors would also like to thank the referees for helpful remarks and suggestions that improved the manuscript.

## References

- 1.Capstick, S. and Keister. B.D.: Multidimensional Quadrature Algorithms at Higher De- gree and/or Dimension, Journal of Computational Physics, 123, 267–273, 1996.Google Scholar
- 2.Chelson, P.: Quasi–random techniques for Monte Carlo methods, PhD Dissertation, The Claremont Graduate School, 1976.Google Scholar
- 3.Hammersley, J. M. and Handscomb, D. C.: Monte Carlo Methods, Methuen, 1964.Google Scholar
- 4.Hickernell, F. J., Hong, H. S.: Computing multivariate normal probabilities using rank-1 lattice sequences, in Proceedings of the Workshop on Scientific Computing, Hong Kong, 1997, G. H. Golub, S. H. Lui, F. T. Luk, and R. J. Plemmons, eds., Springer-Verlag, Singapore, 1997, pp. 209–215Google Scholar
- 5.Hickernell, F.J., Hong, H.S., L’Ecuyer, P., Lemieux, C.: Extensible Lattice Sequences for Quasi-Monte Carlo Quadrature, SIAM J. Sci. Comp., 22, (2000), pp. 1117–1138.Google Scholar
- 6.Hickernell, F.J. and Niederreiter, H.: The existence of good extensible rank-1 lattices, J. Complex., 19, (2003), pp. 286–300.Google Scholar
- 7.Hua, L. K., Wang, Y.: Applications of Number Theory to Numerical Analysis, Springer Verlag, Berlin, 1981.Google Scholar
- 8.Korobov, N. M.: Computation of multiple integrals by the method of optimal coefficients. Vestnik Muskov. Univ. Ser. Mat. Meh. Astr. Fiz. Him., no. 4 (1959), pp. 19–25 (Russian).Google Scholar
- 9.Korobov, N. M.: Number–Theoretic Methods in Approximate Analysis, Fizmatig, Moscow, 1963, (Russian).Google Scholar
- 10.Kuipers, L. and Niederreiter, H.: Uniform Distribution of Sequences, Wiley, New York, 1974.Google Scholar
- 11.Maize, E. Contributions to the theory of error reduction in quasi–Monte Carlo methods, PhD Dissertation, The Claremont Graduate School, 1981.Google Scholar
- 12.Niederreiter, H.: Point sets and sequences with small discrepancy, Monatsch. Math., 104 (1987), pp. 273–337.Google Scholar
- 13.Niederreiter, H.: Random Number Generation and Quasi-Monte Carlo Methods. CBMS-NSF Series in Applied Mathematics, vol. 63. SIAM, Philadelphia, 1992.Google Scholar
- 14.Niederreiter, H. and Pillichshammer, F.: Construction Algorithms for Good Extensible Lattice Rules, Constr. Approx., 30, (2009), pp. 361–393.Google Scholar
- 15.Papageorgiou, A.: Fast Convergence of Quasi-Monte Carlo for a Class of Isotropic Integrals, Mathematics of Computation, 70, Number 233, pp. 297–306, 2000.Google Scholar
- 16.Powell, MJD, Swann, J.: Weighted uniform sampling – a Monte Carlo technique for reducing variance, J. Inst. Maths. Applica., 2 (1966), pp. 228–236.Google Scholar
- 17.Soboĺ, I.M.: The distribution of points in a cube and the approximate evaluation of integrals, Z. Vycisl Mat.i. Mat Fiz., 7, 784–802 = USSR Computational Math and Math Phys., 7 (1967), pp. 86–112.Google Scholar
- 18.Spanier, J., A new family of estimators for random walk problems, J. Inst. Maths. Applica., 23 (1979), pp. 1–31.Google Scholar
- 19.Spanier, J. and Gelbard, E.M., Monte Carlo Principles and Neutron Transport Problems, Addison–Wesley Pub. Co., Inc., Reading, Mass., 1969Google Scholar
- 20.Spanier, J., Li, L.: Quasi-Monte Carlo Methods for Integral Equations, in Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing, Proc. Conf. at University of Salzburg, Austria, July 9–12, 1996, H. Niederreiter, P. Hellekalek, G. Larcher and P. Zinterhof, eds., Springer Lecture Notes on Statistics #127, 1998Google Scholar
- 21.Spanier, J., Maize, E. H.: Quasi-random methods for estimating integrals using relatively small samples, SIAM Rev., 36 (1994), pp. 18–44.Google Scholar
- 22.Zaremba, S.K., ed.: Applications of Number Theory to Numerical Analysis, Academic Press, New York, 1972.Google Scholar