A conjugate prior of a certain likelihood function \(p\qty(x | \theta)\) is a type of prior where the posterior distribution—
\begin{equation} p\qty(\theta | x) \end{equation}
is the same probability family as the prior \(p\qty(\theta)\).
This means that updating is very easy; if we can’t do this, probabilistic programming!