Ability of the radial basis function approach to extrapolate nuclear mass
Tao Li1,2,(),Haiwan Wei1,Min Liu1,2,Ning Wang1,2,()
1Department of Physics, Guangxi Normal University, Guilin, 541004, China 2Guangxi Key Laboratory of Nuclear Physics and Technology, Guilin, 541004, China
(a) Nuclei in the learning set (green region) and two test sets (blue and gray regions) used to examine the extrapolation ability of the RBF approach in the neutron-rich region. (b) Nuclei in the learning set (green region) and two test sets (blue and gray regions) used to examine the extrapolation ability of the RBF approach for the superheavy nuclei."
Figure 1.
Figure 2.
Root-mean-square (rms) deviations with respect to the evaluation masses from AME2016 and the extrapolation masses of each layer of nuclei in the neutron-rich region. The solid squares and circles represent the prediction results of the model and the Model+RBF for 1–12 extrapolated nuclear layers, respectively. The hollow circles represent the results of the Model+RBF for the extrapolation of 6–12 nuclear layers."
Figure 2.
Figure 3.
Deviations between the evaluation masses from AME2016 and the masses predicted by the WS4 model in the extrapolation neutron-rich region. The reconstruction functions S1 and S2 are obtained from learning set 1 and learning set 2, respectively. Figures 3(a)–(l) show the results for 1–12 extrapolated nuclear layers, respectively."
Figure 3.
Figure 4.
Deviations between the evaluation masses from AME2016 and the masses extrapolated by the DZ31 and HFB27 models and the corresponding improved masses extrapolated by the RBF approach in the superheavy region. The solid and hollow blocks represent the results of the DZ31 model and the DZ31+RBF, respectively. The solid and hollow circles represent the results of the HFB27 model and the HFB27+RBF, respectively."
Figure 4.
Figure 5.
Similar to figure 4, but for the WS4 and FRDM12 models."