Strong and weak approximation of semilinear stochastic evolution equations

From MaRDI portal





This monograph is devoted to the study of semilinear stochastic evolution equations (SEEq) in a Hilbert-space setting and to their strong and weak approximation. These equations are of the form: \[ dX(t) +[AX(t) + f(t,X(t))]dt = g(t,X(t))dW(t) \] for \(0\leq t \leq T\), with \(X(0)=X_0\). Here, \(A\) is a linear operator on the Hilbert space \(H\), \(W\) is an \(H\)-valued Wiener space and \(X\) is an \(H\)-valued stochastic process which is the mild solution of the equation. After an introductory chapter, Chapter 2 provides the definition of a mild solution to an SEEq, existence and uniqueness results, and spatial and temporal regularity properties of the mild solution. In Chapter 3, the spatially semidiscrete approximation and the spatio-temporal discretization of the solution are investigated, and optimal strong estimates are obtained. As usual, the study of weak errors is more demanding. For that purpose, the author's idea is using Malliavin calculus in place of Kolmogorov's equation. Chapter 4 is merely a short review of Malliavin calculus in Hilbert spaces. Chapter 5 uses Malliavin calculus and especially Bismut's integration by parts formula to obtain a representation formula of the weak error for both approximation methods. It is also proved that the weak error of convergence is almost twice the order of strong convergence for the discretization error for the linear heat equation with random inhomogeneities. The final Chapter 6 illustrates the theoretical findings for the spatially semidiscrete approximation through a series of numerical experiments.




Cited in
(only showing first 100 items - show all)








This page was built for publication: Strong and weak approximation of semilinear stochastic evolution equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q371442)