JANA: Jointly Amortized Neural Approximation of Complex Bayesian Models
Abstract
This work proposes ``jointly amortized neural approximation'' (JANA) of intractable likelihood functions and posterior densities arising in Bayesian surrogate modeling and simulation-based inference. We train three complementary networks in an end-to-end fashion: 1) a summary network to compress individual data points, sets, or time series into informative embedding vectors; 2) a posterior network to learn an amortized approximate posterior; and 3) a likelihood network to learn an amortized approximate likelihood. Their interaction opens a new route to amortized marginal likelihood and posterior predictive estimation -- two important ingredients of Bayesian workflows that are often too expensive for standard methods. We benchmark the fidelity of JANA on a variety of simulation models against state-of-the-art Bayesian methods and propose a powerful and interpretable diagnostic for joint calibration. In addition, we investigate the ability of recurrent likelihood networks to emulate complex time series models without resorting to hand-crafted summary statistics.
Cite
@article{arxiv.2302.09125,
title = {JANA: Jointly Amortized Neural Approximation of Complex Bayesian Models},
author = {Stefan T. Radev and Marvin Schmitt and Valentin Pratz and Umberto Picchini and Ullrich Köthe and Paul-Christian Bürkner},
journal= {arXiv preprint arXiv:2302.09125},
year = {2023}
}