Closed luiarthur closed 4 years ago
There are two ways?
2 * \bar{D(\theta)} + D(\bar{\theta})
\bar{D(\theta)} + 0.5 * \var{D(\theta)}
D(\theta) = -2 log p(y|theta)
Currently, I'm doing 1. Should I do 2 instead?
Also see this regarding conditional DIC (for missing data)
There are two ways?
2 * \bar{D(\theta)} + D(\bar{\theta})
\bar{D(\theta)} + 0.5 * \var{D(\theta)}
, whereD(\theta) = -2 log p(y|theta)
Currently, I'm doing 1. Should I do 2 instead?
Also see this regarding conditional DIC (for missing data)