GATSBI: Generative Adversarial Training for Simulation-Based Inference
This addresses the challenge of Bayesian inference for complex simulators in fields like physics and engineering, offering a novel method that is incremental by combining GANs with SBI.
The paper tackles the problem of simulation-based inference (SBI) in high-dimensional models by introducing GATSBI, an adversarial approach that learns implicit posterior distributions without explicit likelihoods, showing it outperforms a state-of-the-art SBI method on a camera optics model and provides well-calibrated estimates in high dimensions.
Simulation-based inference (SBI) refers to statistical inference on stochastic models for which we can generate samples, but not compute likelihoods. Like SBI algorithms, generative adversarial networks (GANs) do not require explicit likelihoods. We study the relationship between SBI and GANs, and introduce GATSBI, an adversarial approach to SBI. GATSBI reformulates the variational objective in an adversarial setting to learn implicit posterior distributions. Inference with GATSBI is amortised across observations, works in high-dimensional posterior spaces and supports implicit priors. We evaluate GATSBI on two SBI benchmark problems and on two high-dimensional simulators. On a model for wave propagation on the surface of a shallow water body, we show that GATSBI can return well-calibrated posterior estimates even in high dimensions. On a model of camera optics, it infers a high-dimensional posterior given an implicit prior, and performs better than a state-of-the-art SBI approach. We also show how GATSBI can be extended to perform sequential posterior estimation to focus on individual observations. Overall, GATSBI opens up opportunities for leveraging advances in GANs to perform Bayesian inference on high-dimensional simulation-based models.