Stochastic Gradient Bayesian Optimal Experimental Designs for Simulation-based Inference
This work addresses the problem of efficient experimental design for researchers using SBI in scientific domains, though it is incremental as it extends existing BOED methods to SBI applications.
The paper tackled the challenge of applying Bayesian Optimal Experimental Design (BOED) to simulation-based inference (SBI) models, which are often non-differentiable, by establishing a connection between ratio-based SBI algorithms and stochastic gradient variational inference, enabling simultaneous optimization of designs and inference functions.
Simulation-based inference (SBI) methods tackle complex scientific models with challenging inverse problems. However, SBI models often face a significant hurdle due to their non-differentiable nature, which hampers the use of gradient-based optimization techniques. Bayesian Optimal Experimental Design (BOED) is a powerful approach that aims to make the most efficient use of experimental resources for improved inferences. While stochastic gradient BOED methods have shown promising results in high-dimensional design problems, they have mostly neglected the integration of BOED with SBI due to the difficult non-differentiable property of many SBI simulators. In this work, we establish a crucial connection between ratio-based SBI inference algorithms and stochastic gradient-based variational inference by leveraging mutual information bounds. This connection allows us to extend BOED to SBI applications, enabling the simultaneous optimization of experimental designs and amortized inference functions. We demonstrate our approach on a simple linear model and offer implementation details for practitioners.