Risk-Aware Goal-Oriented Bayesian Optimal Experimental Design
Abstract
Traditional Bayesian optimal experimental design (OED) selects measurements that best inform a model's parameters. However, such measurements can be suboptimal for downstream predictions. Goal-oriented OED targets the prediction directly. However, the existing goal-oriented criteria value all reductions in predictive uncertainty equally, with no way to prioritize rare, high-consequence outcomes. In this article, we develop a risk-aware framework that composes risk at three levels, each generalizing an ingredient of classical I- and G-optimal design: a deviation measure of the posterior predictive uncertainty (generalizing the predictive variance), a risk measure across the prediction domain (interpolating I-optimal averaging and G-optimal worst-case selection), and a risk measure over datasets (generalizing the expectation). We generate each level from a regret function in the risk quadrangle, so that one triple specifies a practitioner's risk preference. We relax the design to continuous weights on the unit simplex and construct a nested-quadrature estimator that is differentiable in the design variable. This enables solving the optimal design problem with gradient-based methods, avoiding a combinatorial search over candidate designs. For a linear-Gaussian lognormal model and a nonlinear extension, we derive closed-form objectives. These give exact references against which we verify that the estimator converges. We demonstrate this framework for finding optimal sensor placements in an inverse problem governed by an advection-diffusion equation. We find that the risk-aware designs substantially outperform the expected-information-gain baseline, which is statistically indistinguishable from a random allocation.
Related
- Certified Bayesian optimal experimental design: from Expected Information Gain to Signal-to-Noise Ratio
- Mathematical FCEV System Modeling for Energy-Consumption-Oriented Optimal Design of an Ultra-High-Speed Air Compressor Motor
- FairBED: Incorporating Fairness into Bayesian Experimental Design
- Power-posterior implicit sampling for efficient Bayesian experimental design: Application to contaminant source identification in groundwater flow
- TestBED: An Open-Source Toolkit for Bayesian Experimental Design
BibTeX
@misc{jakeman2026riskaware,
title = {{Risk-Aware Goal-Oriented Bayesian Optimal Experimental Design}},
author = {John D. Jakeman and Rebekah White and Bart van Bloemen Waanders and Drew P. Kouri and Alen Alexanderian},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.08023},
eprint = {2609.08023},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.08023},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Risk-Aware Goal-Oriented Bayesian Optimal Experimental Design
AU - John D. Jakeman
AU - Rebekah White
AU - Bart van Bloemen Waanders
AU - Drew P. Kouri
AU - Alen Alexanderian
PY - 2026
DA - 2026/09/07
PB - arXiv
DO - 10.48550/arxiv.2609.08023
UR - https://arxiv.org/abs/2609.08023
AB - Traditional Bayesian optimal experimental design (OED) selects measurements that best inform a model's parameters. However, such measurements can be suboptimal for downstream predictions. Goal-oriented OED targets the prediction directly. However, the existing goal-oriented criteria value all reductions in predictive uncertainty equally, with no way to prioritize rare, high-consequence outcomes. In this article, we develop a risk-aware framework that composes risk at three levels, each generalizing an ingredient of classical I- and G-optimal design: a deviation measure of the posterior predictive uncertainty (generalizing the predictive variance), a risk measure across the prediction domain (interpolating I-optimal averaging and G-optimal worst-case selection), and a risk measure over datasets (generalizing the expectation). We generate each level from a regret function in the risk quadrangle, so that one triple specifies a practitioner's risk preference. We relax the design to continuous weights on the unit simplex and construct a nested-quadrature estimator that is differentiable in the design variable. This enables solving the optimal design problem with gradient-based methods, avoiding a combinatorial search over candidate designs. For a linear-Gaussian lognormal model and a nonlinear extension, we derive closed-form objectives. These give exact references against which we verify that the estimator converges. We demonstrate this framework for finding optimal sensor placements in an inverse problem governed by an advection-diffusion equation. We find that the risk-aware designs substantially outperform the expected-information-gain baseline, which is statistically indistinguishable from a random allocation.
ER -