Skip to content

Milestones

List view

  • Integrate models and appropriate codes into the AI/ML controls ecosystem and deploy in Hall-B.

    Overdue by 6 month(s)
    Due by October 1, 2025
  • Utilize the Polarized Target Group’s test facility to test the model and further refine it.

    Overdue by 8 month(s)
    Due by August 1, 2025
  • Test model against simulation and adjust to optimize performance.

    Overdue by 1 year(s)
    Due by May 1, 2025
  • Identify and train an appropriate model for controlling microwave frequency based on historical data and direct NMR feedback. This should start with a Deep Reinforcement Learning model.

    Overdue by 1 year(s)
    Due by February 1, 2025
  • Implement signal extraction technique for accurate extraction of NMR signal from electronics background.

    Overdue by 1 year(s)
    Due by November 1, 2024
    0/2 issues closed
  • Collect historical waveform data for NMR signal from polarized target and prepare for use in model training.

    Overdue by 1 year(s)
    Due by August 1, 2024
    1/1 issues closed
  • Develop simulation of target polarization behavior based on historical archives that can be used for AI/ML model development. Simulation should generate NMR-like waveforms with background and signal changing in a way similar to what is observed in real data. The simulation should be configurable so it can generate much worse and much better conditions than the nominal. This may be written in any language.

    Overdue by 1 year(s)
    Due by July 4, 2024
    0/3 issues closed
  • Identify and curate appropriate historical data sets of measured polarization. This will include the CLAS12 Run Group C archives.

    Overdue by 2 year(s)
    Due by April 4, 2024
    3/3 issues closed