GRETA Filter Parameter Optimization and Calibration

The Gamma-Ray Energy Tracking Array (GRETA) will be a world-leading gamma-ray spectrometer, used at FRIB and ATLAS to explore the properties of exotic nuclei by reconstructing gamma-rays emitted from excited nuclear states with better resolution and peak-to-total than previous generation arrays. While its predecessors were predominantly calibrated and optimized by human scientists, it is planned for GRETA to automate those tasks by utilizing Bayesian optimization and other machine learning techniques. I will present on work to optimize energy resolution by optimizing specific shaping parameters and exploring various criteria for energy resolution in the reconstructions. In creating an optimization-calibration loop that will feed the best filter parameters to the system and calibrate the output to accurate energy values, the human workload to set up GRETA will be greatly reduced and much more efficient. Additionally it will help scientists better understand how GRETA is functioning and the status of each of the detectors throughout its use.