Alpha half-lives calculation of superheavy nuclei with Qα-value predictions based on the Bayesian neural network approach

In this work, we performed a systematic study of the alpha-decay process by employing the Bayesian neural network approach. The Q(alpha)-value prediction of the ten parameter Duflo-Zuker mass model has been improved from a root-mean-square deviation relative to the experimental data sigma = 0.43 MeV to sigma = 0.122 MeV. This correction brought to light some missing physical aspects in the DZ mass model, so as to identify some magic numbers not present in the original model. By using a phenomenological effective model to deal with alpha decay half-lives, we were able to obtain the half-life values throughout the superheavy elements region. As a main result, we found that the region of greatest stability against the alpha-decay process for superheavy elements is between 106 <= Z <= 110 and 180 <= N <= 184.