We will compare the consequences of assuming a discrete or continuous mean for a Dirichlet process in terms of predictive distributions of a survival time. The result of the comparison is that some standard approximations of a discrete distribution with a continuous one are not neutral when they occurs in a nonparametric or semiparametric Bayesian context with a mixture of Dirichlet processes prior for the unknown distribution function of the data.
Medie continue o discrete per un processo di Dirichlet?
CAROTA, Cinzia
2000-01-01
Abstract
We will compare the consequences of assuming a discrete or continuous mean for a Dirichlet process in terms of predictive distributions of a survival time. The result of the comparison is that some standard approximations of a discrete distribution with a continuous one are not neutral when they occurs in a nonparametric or semiparametric Bayesian context with a mixture of Dirichlet processes prior for the unknown distribution function of the data.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
