Fast, accurate and lightweight sequential simulation-based inference using Gaussian locally linear mappings
This addresses the problem of efficient Bayesian inference for complex models with intractable likelihoods, offering a lightweight solution that is incremental but improves computational efficiency.
The paper tackles the trade-off between accuracy and computational demand in simulation-based inference (SBI) by proposing an alternative using structured mixtures of probability distributions, which achieves accurate posterior inference comparable to state-of-the-art neural network methods, even for multimodal posteriors, while having a much smaller computational footprint.
Bayesian inference for complex models with an intractable likelihood can be tackled using algorithms performing many calls to computer simulators. These approaches are collectively known as "simulation-based inference" (SBI). Recent SBI methods have made use of neural networks (NN) to provide approximate, yet expressive constructs for the unavailable likelihood function and the posterior distribution. However, the trade-off between accuracy and computational demand leaves much space for improvement. In this work, we propose an alternative that provides both approximations to the likelihood and the posterior distribution, using structured mixtures of probability distributions. Our approach produces accurate posterior inference when compared to state-of-the-art NN-based SBI methods, even for multimodal posteriors, while exhibiting a much smaller computational footprint. We illustrate our results on several benchmark models from the SBI literature and on a biological model of the translation kinetics after mRNA transfection.