Preconditioned Neural Posterior Estimation for Likelihood-free Inference

TL;DR

This paper introduces preconditioned neural posterior estimation (PNPE) and its sequential version (PSNPE), combining ABC and neural methods to improve likelihood-free inference accuracy, especially in low-dimensional problems.

Contribution

The paper proposes a novel preconditioning approach that enhances neural posterior estimation by integrating ABC, leading to more accurate posterior inference in likelihood-free settings.

Findings

PNPE and PSNPE outperform traditional NPE and SNPE in various examples.

The methods effectively handle complex models like agent-based tumor growth.

Empirical results show improved posterior accuracy over existing neural SBI methods.

Abstract

Simulation based inference (SBI) methods enable the estimation of posterior distributions when the likelihood function is intractable, but where model simulation is feasible. Popular neural approaches to SBI are the neural posterior estimator (NPE) and its sequential version (SNPE). These methods can outperform statistical SBI approaches such as approximate Bayesian computation (ABC), particularly for relatively small numbers of model simulations. However, we show in this paper that the NPE methods are not guaranteed to be highly accurate, even on problems with low dimension. In such settings the posterior cannot be accurately trained over the prior predictive space, and even the sequential extension remains sub-optimal. To overcome this, we propose preconditioned NPE (PNPE) and its sequential version (PSNPE), which uses a short run of ABC to effectively eliminate regions of parameter space that produce large discrepancy between simulations and data and allow the posterior emulator to be more accurately trained. We present comprehensive empirical evidence that this melding of neural and statistical SBI methods improves performance over a range of examples, including a motivating example involving a complex agent-based model applied to real tumour growth data.