Robust design of optimal experiments considering consecutive re-designs

Abstract

We investigate the problem of robust design of experiments (rDoE) in the context of nonlinear maximum-likelihood parameter estimation. It is assumed that an experimenter designs a series of experiments with the possibility of a re-design after a particular experiment run. We present a novel rDoE approach that uses multi-stage decision making in order to explicitly account for the experiment re-designs. This is an extension to our previous work Gottu Mukkula et al. (2021) whereby we focus on the framework of the exact joint-confidence regions for uncertain model parameters. An over-approximation of the exact joint-confidence region is used for designing robust A-optimal experiments. We compare the presented approach with the standard robustification approaches and report the findings on a simple nonlinear case study.