Nonparametric Bayesian estimation and goodness of fit test

Abstract

We first make a review of prior distributions neutral to the right, and then we get the Bayes rule for the survival function S(t) = 1 - F(t), with quadratic loss, with these prior distributions. We give, after that, the estimator with a special kind of processes neutral to the right, the homogeneous processes. We get in point four the linear Bayes rule and we give there an interpretation of the parameters. We finish with a Bayesian generalization of the Kolmogorov-Smirnov goodness of fit test.