Consider a nonparametric Bayesian estimation problem where the Statistician has decided to use an exponential gamm aprocess prior. This paper deals with the selection problem of the process parameters.
Some algorithm to determine the parameter c from the prior guess and the strength of belief are given.
The case where this last concept changes with time is also studied.
@article{urn:eudml:doc:40060,
title = {On the selection of the parameters of an exponential gamma process prior in Bayesian nonparametric estimation.},
journal = {Q\"uestii\'o},
volume = {9},
year = {1985},
pages = {1-5},
mrnumber = {MR0841739},
language = {en},
url = {http://dml.mathdoc.fr/item/urn:eudml:doc:40060}
}
Morales, Domingo. On the selection of the parameters of an exponential gamma process prior in Bayesian nonparametric estimation.. Qüestiió, Tome 9 (1985) pp. 1-5. http://gdmltest.u-ga.fr/item/urn:eudml:doc:40060/