Bayesian Uncertainty Quantification for Anaerobic Digestion models

TL;DR

This paper introduces VarBUQ, a Bayesian uncertainty quantification method for anaerobic digestion models, balancing flexibility and computational efficiency, and demonstrates its advantages over existing methods through synthetic data benchmarks.

Contribution

The paper presents a novel Bayesian procedure, VarBUQ, specifically designed for anaerobic digestion models, improving uncertainty estimation with a good balance of fit and confidence.

Findings

VarBUQ outperforms Fisher's, bootstrap, and Beale's methods in confidence accuracy.

Prior distribution bias enhances Bayesian method performance.

The 'aduq' Python package facilitates implementation of the proposed method.

Abstract

Uncertainty quantification is critical for ensuring adequate predictive power of computational models used in biology. Focusing on two anaerobic digestion models, this article introduces a novel generalized Bayesian procedure, called VarBUQ, ensuring a correct tradeoff between flexibility and computational cost. A benchmark against three existing methods (Fisher's information, bootstrapping and Beale's criteria) was conducted using synthetic data. This Bayesian procedure offered a good compromise between fitting ability and confidence estimation, while the other methods proved to be repeatedly overconfident. The method's performances notably benefitted from inductive bias brought by the prior distribution, although it requires careful construction. This article advocates for more systematic consideration of uncertainty for anaerobic digestion models and showcases a new, computationally efficient Bayesian method. To facilitate future implementations, a Python package called 'aduq' is made available.