Bayesian Optimization for Automatic Termination of Inelastic Neutron-Scattering Experiments
A new method for automatic termination of inelastic neutron-scattering experiments has been proposed, using Bayesian optimization to determine optimal bin widths and a stopping criterion. The approach addresses the inefficiency of measuring beyond equipment resolution, which wastes valuable beam time. The method integrates bin-width optimization with termination decisions, terminating experiments when optimal bin widths fall below target resolutions. Numerical experiments using real inelastic neutron-scattering data validate the approach, demonstrating its potential to improve experimental efficiency. The work is detailed in an arXiv paper (2603.16946) and represents a significant advancement in automated experimental control for neutron scattering.
Key facts
- The method uses Bayesian optimization for bin-width selection.
- It computes a stopping criterion for online decisions on continuing or terminating experiments.
- Experiments are terminated when optimal bin widths become smaller than target resolutions.
- The approach was validated on real inelastic neutron-scattering data.
- The paper is available on arXiv with identifier 2603.16946.
- The method aims to improve experimental efficiency by avoiding unnecessary measurements.
- It addresses the issue of excessive event data in four-dimensional inelastic neutron-scattering experiments.
- The announcement type is replace-cross.
Entities
Institutions
- arXiv