Efficient, Multimodal, and Derivative-Free Bayesian Inference With Fisher-Rao Gradient Flows
This addresses computational challenges in Bayesian inference for science and engineering applications, offering a systematic solution for expensive forward models, multimodality, and lack of gradients, though it is incremental as it builds on existing gradient flow and Kalman methodologies.
The paper tackles efficient approximate sampling for multimodal probability distributions in Bayesian inference for large-scale inverse problems, proposing the Gaussian Mixture Kalman Inversion (GMKI) method, which achieves rapid convergence and handles derivative-free scenarios, as demonstrated in experiments including a Navier-Stokes initial condition recovery.
In this paper, we study efficient approximate sampling for probability distributions known up to normalization constants. We specifically focus on a problem class arising in Bayesian inference for large-scale inverse problems in science and engineering applications. The computational challenges we address with the proposed methodology are: (i) the need for repeated evaluations of expensive forward models; (ii) the potential existence of multiple modes; and (iii) the fact that gradient of, or adjoint solver for, the forward model might not be feasible. While existing Bayesian inference methods meet some of these challenges individually, we propose a framework that tackles all three systematically. Our approach builds upon the Fisher-Rao gradient flow in probability space, yielding a dynamical system for probability densities that converges towards the target distribution at a uniform exponential rate. This rapid convergence is advantageous for the computational burden outlined in (i). We apply Gaussian mixture approximations with operator splitting techniques to simulate the flow numerically; the resulting approximation can capture multiple modes thus addressing (ii). Furthermore, we employ the Kalman methodology to facilitate a derivative-free update of these Gaussian components and their respective weights, addressing the issue in (iii). The proposed methodology results in an efficient derivative-free sampler flexible enough to handle multi-modal distributions: Gaussian Mixture Kalman Inversion (GMKI). The effectiveness of GMKI is demonstrated both theoretically and numerically in several experiments with multimodal target distributions, including proof-of-concept and two-dimensional examples, as well as a large-scale application: recovering the Navier-Stokes initial condition from solution data at positive times.