1: \begin{abstract}

2: In this paper we introduce a Hilbert space-valued Malliavin calculus for Poisson random measures. 

3: It is solely based on elementary principles from the theory of point processes and basic moment estimates, and thus allows for a simple treatment of the Malliavin operators.

4: The main part of the theory is developed for general Poisson random measures, defined on a $\sigma$-finite measure space, with minimal conditions.

5: The theory is shown to apply to a space-time setting, suitable for studying stochastic partial differential equations. 

6: As an application, we analyze the weak order of convergence of space-time approximations for a class of linear equations with $\alpha$-stable noise, $\alpha\in(1,2)$.

7: For a suitable class of test functions, the weak order of convergence is found to be $\alpha$ times the strong order.

8: %OLD VERSION

9: %In this paper we introduce a Hilbert space valued Malliavin calculus for Poisson random measures. We use established theory from the theory of point processes. It allows a simple treatment of the Malliavin operators and moreover it does not rely on chaos expansions. The main part of the theory is developed for general Poisson random measures, defined on a $\sigma$-finite measure spaces, with minimal conditions. The theory is shown to apply to a space-time setting, suitable for studying stochastic partial differential equations. Applications are presented to weak error analysis of space-time approximations of such equations. Weak convergence rates are proved for linear equations, driven by $\alpha$-stable noise, and to semi-linear equation with additive noise, driven by an $L^2$-martingale without Gaussian part. Regularity, in the Malliavin sense, is proved for semilinear equations with multiplicative noise.

10: \end{abstract}

11: