Bayesian inference for poisson process models with censored data.
Albert Y. Lo · Journal of nonparametric statistics · 1992
Bayesian statistical inference for a nonhomogeneous Poisson point process model with censored data is considered. Weighted gamma process priors are founded to be the natural conjugate priors for the cumulative intensity of this model. Upon deflating the prior, the posterior mean of the cumulative intensity becomes a relative of the well-known Nelson-Aalen estimator. Small- and large-sample approximations to the posterior distribution using the Bayesian bootstrap methods are discussed.