Quasi-Monte Carlo and Multilevel Monte Carlo Methods for Computing Posterior Expectations in Elliptic Inverse Problems
For practitioners solving Bayesian inverse problems with PDEs, this work provides theoretical guarantees that ratio estimation does not increase complexity, enabling efficient uncertainty quantification.
The paper develops a framework for computing posterior expectations in elliptic inverse problems by reducing them to a ratio of prior expectations, and proves that quasi-Monte Carlo and multilevel Monte Carlo methods achieve the same computational complexity for the ratio as for the individual expectations.
We are interested in computing the expectation of a functional of a PDE solution under a Bayesian posterior distribution. Using Bayes' rule, we reduce the problem to estimating the ratio of two related prior expectations. For a model elliptic problem, we provide a full convergence and complexity analysis of the ratio estimator in the case where Monte Carlo, quasi-Monte Carlo or multilevel Monte Carlo methods are used as estimators for the two prior expectations. We show that the computational complexity of the ratio estimator to achieve a given accuracy is the same as the corresponding complexity of the individual estimators for the numerator and the denominator. We {also include numerical simulations, in the context of the model elliptic problem, which demonstrate the effectiveness of the approach.