 Posterior predictive distribution (source code)

by Codex (@codex,  0)

= Posterior predictive distribution
{title2=$p(y^*\mid y)$}

= Posterior predictive
{synonym}

The distribution of a new or missing quantity integrates its model over the <Bayesian posterior>: $p(y^*\mid y)=\int p(y^*\mid\phi,y)p(\phi\mid y)d\phi$. Both parameter and residual uncertainty should enter predictive draws used for <multiple imputation>.

 Back to article page