| Type | Seminar |
| La physique dans tous ses états | |
| Date | June 30, 2026 - 11:00 |
| Time | 11:00 |
| Location | Room 105, GANIL, Caen | France |
by Emmanuel Clément
The seminar will present a predictive model for the first excited states in even-even nuclei based on machine learning (ML). The model was trained using ground state properties and excited states from the evaluated Nuclear Data Evaluation Committee (NNDC) database solely with minimum ingredient in the input layer. The overall agreement is remarkable. The resulting model captures most of the nuclear structure evolution from oxygen to uranium and can predict the features of very exotic nuclei that have not yet been measured experimentally. Several isotopic chains are detailed, including those of Sn, Ni, Zr and Ca.
