Error Estimation for Quasi-Monte Carlo Methods
A hybrid-Monte Carlo method designed to obtain a statistical error analysis using deterministic sequences is introduced. The method, which is called “random sampling from low-discrepancy sequences”, produces estimates that satisfy deterministic error bounds yet confidence interval analysis can be applied to measure the accuracy of them.
A particular implementation of the hybrid method is applied to three problems; one from mathematical finance and others from particle transport theory. The method is compared with the pseudorandom method and two quasirandom methods. Encouraging numerical results in favor of the hybrid method are obtained.
KeywordsSample Standard Deviation Monte Carlo Sequence European Call Option Mixed Sequence Root Sequence
Unable to display preview. Download preview PDF.
- J. Spanier, “Quasi-Monte Carlo Methods for Particle Transport Problems”, Proc. Conf. on Monte Carlo Methods in Scientific Computing, Univ. Las Vegas, 1994.Google Scholar
- B. S. Moskowitz, “Application of quasi-random sequences to Monte Carlo methods”, Ph. D. Dissertation, UCLA, 1993.Google Scholar
- A. B. Owen, “Randomly Permuted (t, m, s)—Nets and (t,s)—Sequences”, Proc. Conf. on Monte Carlo Methods in Scientific Computing, Univ. Las Vegas, 1994.Google Scholar
- J. C. Hull, Options,Futures, And Other Derivative Securities, Prentice-Hall, Inc., 1989.Google Scholar
- G. Pagès and Y. J. Xiao, “Sequences with low discrepancy and pseudorandom numbers: theoretical remarks and numerical tests”, Technical Report, Laboratoire de Mathematiques et Modelisation, Paris, 1991.Google Scholar