Machine Learning Based Classification of the Halos in Light Nuclei Region

Abstract

Experimental and theoretical studies on halo nuclei, whose nucleon binding energies are extremely weak, are among the most interesting topics of nuclear physics studies. By better defining and understanding this unusual behavior of these nuclei, our understanding of nuclear structure can be further improved. Although there are already a few experimentally proven halo nuclei in the literature, many others have found their place in the literature as candidate halo nuclei. In this study, the classification of halo nuclei was carried out using an artificial neural network approach. In the light nuclei region, the properties of nuclei, including halo nuclei, were discussed and the existing halo nuclei were classified. The success of the obtained results indicates that machine learning methods can be used for identifying halo nuclei. Thus, these methods are considered as one of the alternative tools to confirm the existence of new or candidate halo nuclei.