Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5127987 | Mathematics and Computers in Simulation | 2018 | 15 Pages |
The simulation of the expectation of a stochastic quantity E[Y] by Monte Carlo methods is known to be computationally expensive especially if the stochastic quantity or its approximation Yn is expensive to simulate, e.g., the solution of a stochastic partial differential equation. If the convergence of Yn to Y in terms of the error |E[YâYn]| is to be simulated, this will typically be done by a Monte Carlo method, i.e., |E[Y]âEN[Yn]| is computed. In this article upper and lower bounds for the additional error caused by this are determined and compared to those of |EN[YâYn]|, which are found to be smaller. Furthermore, the corresponding results for multilevel Monte Carlo estimators, for which the additional sampling error converges with the same rate as |E[YâYn]|, are presented. Simulations of a stochastic heat equation driven by multiplicative Wiener noise and a geometric Brownian motion are performed which confirm the theoretical results and show the consequences of the presented theory for weak error simulations.