Speaker
Martha Papadopoulou
Description
This study investigates the prediction of Total Kinetic Energy in neutron-induced fission using Gaussian Process Regression, a non-parametric method for probabilistic regression. Due to the computational cost of theoretical calculations and the limited availability of experimental data, it is of interest to investigate machine learning methods. Training data are obtained from the General Description of Fission Observables simulation. The GPR model is trained on selected isotopes and incident energies and evaluated on unseen cases. The results show that the model captures the overall behavior of TKE within the range of the training features.
Author
Co-authors
Prof.
Vaia Prassa
(University of Thessaly)
Charalampos Moustakidis
(Department of Physics Aristotle University of Thessaloniki)