ISIS Neutron and Muon Source Data Journal

Neural Network Reconstruction of Low-Statistics ?SR Spectra to Maximize Beamline Efficiency

Abstract: Neural Network is a powerful model to handle big MuSR data. Low-statistics spectra often arise when beamtime is limited or when the beam current/status fluctuates. Therefore, we develop an AI-MuSR framework that uses Multilayer Perceptrons and Kolmogorov Arnold Network regressors to reconstruct high-statistics spectra from low-statistics inputs in real time; assuming the instrument noise characteristics are consistent across materials. To enrich and validate training, we will measure high-purity copper to obtain clean reference field-dependent spectra in event mode; high muon event count runs will be sliced to form matched low-/high-quality pairs. Crucially, if the model sees only temperature sweeps during training, it may mistake field-driven features for noise, so we will include systematic field-dependent measurements with particular emphasis on TF measurements.

Principal Investigator: Dr Muhammad Rabie Bin Omar
Experimenter: Mr KEN FUI CHIN
Experimenter: Professor Takayuki Goto
Local Contact: Dr James Lord
Experimenter: Dr Wan Nurfadhilah Binti Zaharim
Experimenter: Dr Isao Watanabe
Experimenter: Dr Ahmad Rujhan Bin Mohd Rais

DOI: 10.5286/ISIS.E.RB2610300

ISIS Experiment Number: RB2610300

Publisher: STFC ISIS Neutron and Muon Source

Data format: RAW/Nexus
Select the data format above to find out more about it.

Data Citation

The recommended format for citing this dataset in a research publication is as:
[author], [date], [title], [publisher], [doi]

For Example:
Dr Muhammad Rabie Bin Omar et al; (2026): Neural Network Reconstruction of Low-Statistics ?SR Spectra to Maximize Beamline Efficiency, STFC ISIS Neutron and Muon Source, https://doi.org/10.5286/ISIS.E.RB2610300

Data is released under the CC-BY-4.0 license.