Systematic Study on the α-particle preformation factor in the theory of α-decay based on the Tabular Prior-data Fitted Network (TabPFN)

2025/11/18 by Qi, Panpan, Xiao, Xuanpeng, Yu, Gongming +2
#FOS: Physical sciences #Nuclear Theory (nucl-th)

paper · doi:10.48550/arxiv.2511.14705

Abstract

A hybrid approach combining the Tabular Prior-data Fitted Network (TabPFN) with the Coulomb and Proximity Potential Model (CPPM) is developed to investigate α-particle preformation factors P_α and their impact on α-decay half-lives. The TabPFN model, trained on 498 nuclei, accurately learns the relationship between the properties of the nuclear structure and P_α, achieving a root mean square deviation of σ_rms = 0.211. The predicted factors reveal clear odd-even staggering and shell closure effects, and exhibit a linear correlation with Q_α^-1/2, extending the Geiger-Nuttall systematics. When incorporated into CPPM calculations, the machine learning-based P_α values significantly improve half-life predictions. The capability of the model is demonstrated through predictions for superheavy nuclei (Z = 117--120), suggesting N = 184 as a potential neutron magic number.

Citations

• Systematic calculations of cluster radioactivity half-lives in trans-lead nuclei
• TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second
• Machine learning and the physical sciences
• Deep Learning in Neural Networks: An Overview
• A Closer Look at TabPFN v2: Understanding Its Strengths and Extending Its Capabilities

Related

• Bayesian optimization and nonlocal effects method for α decay of superheavy nuclei based on CPPM
• A global study of α-clusters decay in heavy and superheavy nuclei with half-life and preformation factor
• Machine Learning-Driven High-Precision Model for α-Decay Energy and Half-Life Prediction of superheavy nuclei
• Study on nuclear α-decay energy by an artificial neural network with pairing and shell effects
• Non-negligible influence of shape inheritance and staggering on α decay