Schematic figure of the analysis pathway that informs nuclear physics with differential and integral information (Neudecker et al, Phys. Rev. C 104, 034611)

ML/AI methods have been introduced in nuclear reactor applications during the last decades; they contribute today to enhance the safety of the current fleet of reactors.
however, these statistic techniques are becoming increasingly more practical and powerful in recent years. Hence, these current developments foretell an increase of applying ML/AI-based methods to solving issues pertaining to nuclear science and technology.

Professor Óscar Cabellos is coordinating efforts on Machine Learning in the JEFF project. JEFF is the Joint Evaluated Fission and Fusion project which is a collaboration project between NEA Data Bank and participating countries to develop a reference nuclear data library. Technical sessions on ML/AI are scheduled within the JEFF Nuclear Data Weeks. Prof. Cabellos collaborates with experts in ML techniques to study nuclear data integrating differential and integral experiments. The combination of information from differential experimental data, nuclear-physics theory, integral experiments, and neutron-transport simulations of the latter experiments with the help of the random forest algorithm and expert judgment allows to improve the description of nuclear-physics observables as applied to a particular application area.