Compositional simulation-based inference for time series
This work addresses a bottleneck in Bayesian inference for time series data in fields like ecology and epidemiology, offering a more efficient approach but is incremental in its method adaptation.
The paper tackles the challenge of scaling amortized simulation-based inference (SBI) to time series data by proposing a method that exploits Markovian simulators to identify parameters locally from individual state transitions and composes them for global inference, demonstrating improved simulation efficiency on synthetic benchmarks and high-dimensional simulators.
Amortized simulation-based inference (SBI) methods train neural networks on simulated data to perform Bayesian inference. While this strategy avoids the need for tractable likelihoods, it often requires a large number of simulations and has been challenging to scale to time series data. Scientific simulators frequently emulate real-world dynamics through thousands of single-state transitions over time. We propose an SBI approach that can exploit such Markovian simulators by locally identifying parameters consistent with individual state transitions. We then compose these local results to obtain a posterior over parameters that align with the entire time series observation. We focus on applying this approach to neural posterior score estimation but also show how it can be applied, e.g., to neural likelihood (ratio) estimation. We demonstrate that our approach is more simulation-efficient than directly estimating the global posterior on several synthetic benchmark tasks and simulators used in ecology and epidemiology. Finally, we validate scalability and simulation efficiency of our approach by applying it to a high-dimensional Kolmogorov flow simulator with around one million data dimensions.