Marginalization
Gist
If we're looking at a Distribution - Joint , marginalization looks at the distribution when we only look at a subset of variables. Specifically, what would the distribution of X be if I ignored Y?
Properties
- If
and are discrete: