Bayesian Analysis of Change Point Problems Using Conditionally Specified Priors

Abstract In data analysis, change point problems correspond to abrupt changes in stochastic mechanisms generating data. The detection of points is a relevant problem the analysis and prediction time series. this paper, we consider class conjugate prior distributions obtained from conditional specification methodology for solving problem. We illustrate application such Bayesian with Poisson processes. obtain posterior distribution model parameters using general bivariate gamma conditionals. Simulation are readily implemented Gibbs sampling algorithm. even when densities that incompatible or only compatible an improper joint density. methods will be demonstrated examples simulated real