Physics-informed machine learning for radioisotope identification

Machine learning methods in gamma spectroscopy have demonstrated an impressive ability to provide accurate, real-time classification of unknown radioactive samples. However, obtaining sufficient experimental training data is often prohibitively expensive and time-consuming, and models trained solely on synthetic data can struggle to generalize to the unpredictable range of experimental operating scenarios. In this work, we pretrain a model using physically derived synthetic data and subsequently leverage transfer learning techniques to fine-tune the model for a specific target domain. This paradigm enables us to embed physical principles during the pretraining step, thus requiring less data from the target domain compared to existing machine learning methods. The results of this study serve as proof of concept for applying physics-informed machine learning models to scenarios where access to experimental data is limited.