Machine Learning and Artificial Intelligence for computational nuclear data
IAEA: AI 4 Atoms
Invited presentation on 10/2021
Machine Learning and Artificial Intelligence methods offer wide applicability to a class of problems know collectively as "inverse problems." The solution to such problems involves the calculation of the causal factors that are responsible for a set of observations. Inverse problems are frequently encountered in nuclear data, where measurements exist and these observations must be reconciled with model interpretations. We explore probabilistic machine learning techniques in the context of binding energy of atomic nuclei. The set of observations comes from the Atomic Mass Evaluation (AME), which totals over 2000 data points. We explore a rich possibility of physical input spaces and construct neural networks that are capable of describing these observations. Because our method is stochastic, it also provides an estimate of uncertainty for each prediction. We discuss the capacity of such modeling to extrapolate into unmeasured nuclei.
LA-UR-21-30576
Related Publications
| Year | Authors | Title (Click for more details) | Journal (PDF) |
|---|---|---|---|
| 2022 | A. Lovell, A. T. Mohan, T. M. Sprouse, M. Mumpower | Nuclear masses learned from a probabilistic neural network | PRC 106 014305 |