“…Bayesian inference utilizes a second-order random walk to smooth the prior effects of age, period, and cohort [ [39] , [40] , [41] ]. It has been observed to provide better coverage and accuracy compared to traditional projection models such as the Nordpred model, smooth spline model, and Poisson regression [ 24 , 38 , [42] , [43] , [44] ]. To evaluate the performance of BAPC models, the data from 1990 to 2019 were divided into two intervals (1990–2014 and 2015–2019) [ 45 ].…”