Optimal Experimental Design

Formalization of 'how best to acquire data' — criteria and computation for choosing experiments, surveyed in Acta Numerica 2024.

Specifications & Metrics

Optimal experimental design (OED) formalizes questions of how best to acquire data and creates computational methods to answer them.
The Bayesian and decision-theoretic approach is flexible, encompassing information-based criteria well-suited to nonlinear and non-Gaussian models.
The 2024 Acta Numerica survey spans foundations in classical design theory through current research involving OED for complex models.