CARIBU-matic and the MUSIC ML project: examples of machine-learning applications for beam tuning and experimental data analysis/classification

During experiments at accelerator facilities, timely production and transport of the requested beam to the users target station is crucial for both the experiment success and to maintain high user satisfaction. During the data collection phase of an experiment, researchers need fast and accurate methods to process and classify the data produced by their detector systems to verify if changes to the experimental conditions need to be carried out.

Beam tuning and experimental data analysis (or classification) are examples of time sensitive activities done at accelerator facilities. However, typical beam tuning and data classification processes are based on expert-driven manual methods which can be quite time consuming and inefficient. These deficiencies could negatively impact the outcome of an experiment. In this presentation we will discuss two separate projects that leverage machine-learning (ML) methods to optimize specific tasks in the tuning of radioactive beams from the CARIBU facility, and in the classification of data collected with the MUSIC detector.

The 1st project, CARIBU-matic, aims to develop an automated beam tuning system for the CARIBU facility, located inside the Argonne Tandem Linac Accelerator System (ATLAS), by integrating existing diagnostic detectors with a Bayesian optimization machine learning algorithm.

The 2nd project, MUSIC-ML, focuses on the strip-wise classification of specific α-induced reactions within the MUSIC detector. By combining statistical and machine learning methods, we have successfully developed a novel method for identifying these reactions [1]. This new method has been applied to two data sets produced by experiments done at the ATLAS accelerator facility involving the 17F(α,p) and 17O(α,n) reactions. Furthermore, our research demonstrates that the newly developed method outperforms several out-of-the-box techniques for outlier detection.