This is a simple survey of the main methods used in simulation in the aim of computing expectations of random variables : - The Monte Carlo method, based on the large number theorem and pseudo-random sequences, - The quasi-Monte Carlo methods, based on low discrepancy sequences, - Hybrid methods, - the shift method, based on the pointwise ergodic theorem. The case of infinite dimensional settings is also discussed.
Publié le : 1993-01-05
Classification:
pseudo-random generator,
itareted logarithm,
equidistribution,
equidistributed,
translation of the torus,
stoping time,
Wiener space,
[MATH.MATH-NA]Mathematics [math]/Numerical Analysis [math.NA]
@article{hal-00451820,
author = {Bouleau, Nicolas},
title = {Calculs d'esp\'erance par simulation},
journal = {HAL},
volume = {1993},
number = {0},
year = {1993},
language = {fr},
url = {http://dml.mathdoc.fr/item/hal-00451820}
}
Bouleau, Nicolas. Calculs d'espérance par simulation. HAL, Tome 1993 (1993) no. 0, . http://gdmltest.u-ga.fr/item/hal-00451820/