Bayesian Symbolic Regression for Missing Physics

Arno Strouwen

stat.MLMar 16, 2026

Bayesian Symbolic Regression for Missing Physics

Authors: Arno Strouwen

Organizations: Strouwen Statistics · Biosystems Department, KULeuven, Leuven, Belgium

Abstract

Model-based approaches for (bio)process systems often suffer from incomplete knowledge of the underlying physical, chemical, or biological laws. Universal differential equations, which embed neural networks within differential equations, have emerged as powerful tools to learn this missing physics from experimental data. However, neural networks are inherently opaque, motivating their post-processing via symbolic regression to obtain interpretable mathematical expressions. Genetic algorithm-based symbolic regression is a popular approach for this post-processing step, but provides only point estimates and cannot quantify the confidence we should place in a discovered equation. We address this limitation by applying Bayesian symbolic regression, which uses Reversible Jump Markov Chain Monte Carlo to sample from the posterior distribution over symbolic expression trees. This approach naturally quantifies uncertainty in the recovered model structure. We demonstrate the methodology on a Lotka-Volterra predator-prey system and then show how a well-designed experiment leads to lower uncertainty in a fed-batch bioreactor case study.

Figures & tables

Term	Expression	Chains
$\phi_{1}$	$-(x_{1}x_{2})$	5
(true: $-0.9x_{1}x_{2}$ )	$(x_{1}^{2}x_{2})/-(x_{1})$	1
	$-(x_{1})/(x_{1}/(x_{1}x_{2}))$	1
	$x_{1}+((x_{2}x_{1}+x_{1})/(-(x_{2})/x_{2}))$	1
	$-\lt(x_{2}x_{1},0.95,-0.18)$	1
	$\lt(x_{2},0.35,-0.46)+x_{1}^{2}/(((x_{1}/x_{2})/-(x_{1}))\cdot x_{1})$	1

Table 1: Posterior samples from Bayesian symbolic regression for the Lotka-Volterra missing physics. Here

\lt(x,a,b)=ax+b

Figure 1: Lotka-Volterra system. Solid lines: true state values. Thick dots: noisy measurements of states. Dashed lines: predicted states after replacing missing physics with a neural network. Dotted lines: predicted states after replacing missing physics with 10 draws from Bayesian symbolic regression.

Panel of Figure 2 — Figure 2: Expression tree representing $(2x_{1}+1)+\sin(x_{2})$ . Circular nodes contain operators ( $+$ , $\lt$ , $\sin$ ). Rectangular nodes are terminals: rounded rectangles denote features ( $x_{1}$ , $x_{2}$ ), and plain rectangles denote constants. The $\lt$ node implements the linear transformation $\lt(x_{1},2,1)=2x_{1}+1$ .

Figure 3: Bioreactor system. States and measurements.

	Simplified form	$\hat{\mu}_{\max}$	$\hat{K}_{s}$
Opt. 1	$C_{s}/(2.33\,C_{s}+11.09)$	0.43	4.77
Opt. 2	non-polynomial ( $\exp(-C_{s})$ terms)	—	—
Opt. 3	$C_{s}/(2.43\,C_{s}+9.41)$	0.41	3.88
Opt. 4	$C_{s}/(2.38\,C_{s}+10.83)$	0.42	4.55
Opt. 5	$C_{s}/(2.25\,C_{s}+12.44)$	0.44	5.53
True	$\mu_{\max}\,C_{s}/(K_{s}+C_{s})$	0.421	4.39

Table 2: Algebraically simplified equations for the bioreactor with estimated Monod parameters. The optimal experiment recovers the Monod form, while the random experiment does not.

Panel of Figure 4 — Figure 4: Fed-batch bioreactor system. Blue solid line: missing physics, the Monod equation. Thick orange dots: predictions of the missing physics by the neural network. Green dashed lines: predicted missing physics by the top 10 trees generated by Bayesian symbolic regression.

Explore similar work

Bayesian Symbolic Regression with Entropic Reinforcement Learning
Aug 10, 2026Oussama Boussif, Mohammed Mahfoud, Younesse Kaddar +6Bayesian Inference Symbolic Regression
Mila – Québec AI Institute · Université de Montréal · Independent +5
Experimental Design for Missing Physics
Mar 21, 2026Arno Strouwen, Sebastián Micluţa-Câmpeanu
JuliaHub, MA 02111-1929 USA · Biosystems Department, KULeuven, Leuven, Belgium · Faculty of Physics, University of Bucharest, Bucharest, Romania
Deep Variational Inference Symbolic Regression
May 1, 2026James Butterworth, Gevik Grigorian, Alejandro DiazDelaOUncertainty Quantification Variational Inference
Clinical Operational Research Unit University College London London, United Kingdom · Department of Computer Science University College London London, United Kingdom

Aug 10, 2026J/K move · Enter open · S save

Bayesian Symbolic Regression with Entropic Reinforcement Learning

Oussama Boussif, Mohammed Mahfoud, Younesse Kaddar, Moksh Jain+5

Mila – Québec AI Institute · Université de Montréal · Independent+5

Bayesian Inference Symbolic Regression

Open paper AskPDF arXiv

Symbolic regression is the problem of finding an algebraic expression describing a stochastic dependence of a target variable on a set of inputs. Unlike forms of regression that fit parameters assuming a fixed model structure, symbolic regression is a search problem over the space of expressions, represented, for example, as abstract syntax trees using a library of operators. Symbolic regression is typically used in settings with limited, noisy data in the natural sciences. However, searching for a single best-fitting expression fails to capture the epistemic uncertainty about the expression, which motivates a Bayesian perspective that enables uncertainty quantification and specification of natural priors to constrain the search space. In this work, we propose ERRLESS (Entropy-Regularized Reinforcement Learning for Expression Structure Sampling), a scalable approach for sampling from the posterior distribution over expressions given data using maximum-entropy reinforcement learning. ERRLESS learns a neural policy that constructs expressions sequentially by building up their abstract syntax trees. At convergence, the policy samples expressions from the posterior. At test time, expressions can be sampled by rollouts of this policy. We demonstrate that ERRLESS achieves competitive results on the Feynman benchmark while producing short and interpretable expressions. Additionally, we demonstrate that the mean of the posterior predictive approximated by ERRLESS achieves a high coefficient of determination (

R^2

) compared to an SMC baseline, highlighting the benefits of the Bayesian perspective in symbolic regression.

Abstract

Explore similar work

Bayesian Symbolic Regression with Entropic Reinforcement Learning

Experimental Design for Missing Physics

Deep Variational Inference Symbolic Regression