A Machine-learning method for spin classification of neutron resonances

Neutron resonances are sharp fluctuations seen in neutron transmission and capture experiments at low-energy neutron-induced reactions. Properties of neutron resonances are some of the few experimental constraints to nuclear level densities and gamma strength functions (crucial for modeling many nuclear applications). Resonances are characterized by their angular momenta quantum numbers, which are normally assigned through fits often done in an ad hoc and not fully-reproducible manner. Comprehensive compilations of evaluated resonances often contain incorrectly-assigned spins. To address these, we developed a Machine-Learning method to train an algorithm to identify resonances with incorrect spin assignments. Model training is done on synthetic data constructed to simulate statistical properties of resonances seen in real nuclei. The trained classifier can be applied to resonances sequences from compiled, evaluated, or experimental data. In this work we use ⁵²Cr as a test case to assess the performance of the reclassification, showing how multiple realizations of synthetic data can serve as a validation tool for the machine-learning classifier. We then apply the trained algorithm to make reclassification predictions on a ⁵²Cr evaluated file.