[3I04-08] Fission Reaction

[3I06] Analysis of fission trajectories based on machine learning

*Chikako Ishizuka¹, Yuta Mukobara¹, Satoshi Chiba^1,2 (1. Tokyo Tech, 2. NAT)

Keywords:Nuclear fission, Nuclear data, Machine learning

The 4D Langevin model developed by us can comprehensively explain the nature of nuclear fission. Due to this capability, we are undertaking the challenge of extracting new insights into the nuclear fission mechanism by combining Langevin model nuclear fission trajectories with machine learning. In the previous presentation, we reported on the predictive accuracy of a machine learning model trained on the nuclear fission trajectories of the 4D Langevin model. In this presentation, we will report the results of feature analysis from neural network weights applied to this machine learning model, in conjunction with traditional nuclear fission trajectory analysis.