The Gaussian distribution has ever been the most popular and usable device in the field of statistics. Even in the context of penalised complexity (PC) priors, the normal density has a particular meaning, especially because we can consider it as a base model which could be extended both in terms of tail thickness and skewness. We derive the numerical PC prior for the shape parameter of the skew-normal density and the analytical PC prior for the degrees of freedom of the t-distribution. We also perform an approximation of the Kullback-Leibler divergence (KLD) in the the skew-normal model.
Penalising the complexity of extensions of the Gaussian distribution
diego battagliese
Writing – Original Draft Preparation
;
2020-01-01
Abstract
The Gaussian distribution has ever been the most popular and usable device in the field of statistics. Even in the context of penalised complexity (PC) priors, the normal density has a particular meaning, especially because we can consider it as a base model which could be extended both in terms of tail thickness and skewness. We derive the numerical PC prior for the shape parameter of the skew-normal density and the analytical PC prior for the degrees of freedom of the t-distribution. We also perform an approximation of the Kullback-Leibler divergence (KLD) in the the skew-normal model.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.
