摘要：In this work we perform a Bayesian inference of the crust-core transition density ρt of neutron stars based upon the neutron-star radius and the neutron-skin thickness data within a thermodynamical method. The uniform and Gaussian distributions for the ρt prior are adopted in the Bayesian approach. It has a larger probability to have values higher than 0.1 fm 3 for ρt as the uniform prior and the neutron-star radius data are used. We found that this is controlled by the curvature Ksym of the nuclear symmetry energy. This phenomenon will not happen if Ksym is not extremely negative, i.e., Ksym > -200 MeV. The obtained ρt is 0.075+0.005-0.01fm^3 at 68% confifidence level when both the neutron-star radius and the neutron-skin thickness data are taken into account. The strongly anti-correlations between ρt and the slope L, curvature of the nuclear symmetry energy are observed. The dependence of the three L-Ksym correlations predicted in the literature on the crust-core density and pressure is quantitatively investigated. The most probable value of 0.08 fm^3 for ρt is obtained from the L-Ksym relation raised by Holt et al. and the larger values are preferred by the other two relations.