Prediction of Nuclear Charge Density Distribution with Feedback Neural Network

Abstract

The nuclear charge density distribution plays an important role in nuclear physics and atomic physics. As one of the most frequently used models to obtain charge density distribution, the two-parameter fermi (2pF) model has been widely applied in both nuclear physics and atomic physics. Currently, the feedforward neural network has been employed to study the available 2pF model parameters for 86 nuclei, and it is found that by introducing A^{1/3} into the input parameter of the neural network, the accuracy and precision of the parameter learning effect are improved. Furthermore, the average result of multiple predictions is more reliable than the best result of a single prediction, and there is no significant difference between the average result of density value and of parameter value for the average charge density distribution. In addition, 2pF parameters of 284 (near) stable nuclei are also predicted in this work, which provides a reference for the experiment.