Presentation
23 August 2023 Efficient reconstruction of model parameters using Bayesian target-vector optimization
Matthias Plock, Sven Burger, Philipp-Immanuel Schneider
Author Affiliations +
Abstract
Parameter reconstruction problems appear frequently in optical metrology. Here, one attempts to explain a set of K experimental measurements by fitting to them a parameterized forward model of the measurement process. We present a Bayesian target vector optimization scheme that can be used to perform this fit. It has been shown to be capable of outperforming established methods such as Levenberg-Marquardt, and can after a successful fit enable very efficient and accurate determination of the distribution of the reconstructed model parameters using Markov chain Monte Carlo sampling.
Conference Presentation
© (2023) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Matthias Plock, Sven Burger, and Philipp-Immanuel Schneider "Efficient reconstruction of model parameters using Bayesian target-vector optimization", Proc. SPIE PC12619, Modeling Aspects in Optical Metrology IX, PC1261905 (23 August 2023); https://doi.org/10.1117/12.2673590
Advertisement
Advertisement
KEYWORDS
Mathematical optimization

Education and training

Monte Carlo methods

Optical metrology

Process modeling

Computer programming languages

Global Positioning System

Back to Top