Rare event simulation via importance sampling for linear SPDE's (Q1706675)

The current work is the first one (to the best knowledge of the authors) to address provably efficient importance sampling schemes for SPDEs that perform well both in the limit as the noise goes to zero, but in the prelimit as well, and that do not degrade in performance due to increased dimension. Efficient importance sampling Monte Carlo methods for the estimation of rare events within the class of linear stochastic partial differential equations are given. If a spectral gap of appropriate size exists, then one can identify a lower dimensional manifold where the rare event takes place. The theoretical results of this paper are demonstrated by a series of simulation studies.