Likelihood-Free Parameter Estimation with Neural Bayes Estimators

Abstract

Neural point estimators are neural networks that map data to parameter point estimates. They are fast, likelihood free and, due to their amortised nature, amenable to fast bootstrap-based uncertainty quantification. In this paper, we aim to increase the awareness of statisticians to this relatively new inferential tool, and to facilitate its adoption by providing user-friendly open-source software. We also give attention to the ubiquitous problem of making inference from replicated data, which we address in the neural setting using permutation-invariant neural networks. Through extensive simulation studies we show that these neural point estimators can quickly and optimally (in a Bayes sense) estimate parameters in weakly-identified and highly-parameterised models with relative ease. We demonstrate their applicability through an analysis of extreme sea-surface temperature in the Red Sea where, after training, we obtain parameter estimates and bootstrap-based confidence intervals from hundreds of spatial fields in a fraction of a second.

The American Statistician (2023)

References in corpus (4)

• Deep Integro-Difference Equation Models for Spatio-Temporal Forecasting
• Neural parameter calibration for large-scale multi-agent models
• Vecchia Likelihood Approximation for Accurate and Fast Inference in Intractable Spatial Extremes Models
• Likelihood-Free Inference with Generative Neural Networks via Scoring Rule Minimization

Cited by in corpus (8)

• Neural Likelihood Surfaces for Spatial Processes with Computationally Intensive or Intractable Likelihoods
• A picture guide to cancer progression and monotonic accumulation models: evolutionary assumptions, plausible interpretations, and alternative uses
• Flexible space-time models for extreme data
• Latent Variable Sequence Identification for Cognitive Models with Neural Network Estimators
• Deep Optimal Experimental Design for Parameter Estimation Problems
• Regularization of the ensemble Kalman filter using a non-parametric, non-stationary spatial model
• Neural Bayes estimation and selection for complex bivariate extremal dependence models
• Trustworthy scientific inference with generative models