Index: The Book of Statistical ProofsGeneral TheoremsBayesian statisticsProbabilistic modeling ▷ Generative model

Definition: Consider measured data $y$ and some unknown latent parameters $\theta$. A statement about the distribution of $y$ given $\theta$ is called a generative model $m$

\[\label{eq:gm} m: \, y \sim \mathcal{D}(\theta) \; ,\]

where $\mathcal{D}$ denotes an arbitrary probability distribution and $\theta$ are the parameters of this distribution.

 
Sources:

Metadata: ID: D27 | shortcut: gm | author: JoramSoch | date: 2020-03-03, 15:50.