- #1
jimmy1
- 61
- 0
A probability distribution,[tex]f(x) [/tex] ,can be represented as a generating function,[tex]G(n) [/tex], as [tex] \sum_{x} f(x) n^x [/tex]. The expectation of [tex]f(x) [/tex] can be got from [tex] G'(1) [/tex].
A bivariate generating function, [tex]G(m,n) [/tex] of the joint distribution [tex] f(x,y) [/tex] can be represented as [tex] \sum_{x} \sum_{y} f(x,y) n^x m^y [/tex].
Now my question is how can I get the expectation of [tex] f(x,y) [/tex] from the above generating function?
A bivariate generating function, [tex]G(m,n) [/tex] of the joint distribution [tex] f(x,y) [/tex] can be represented as [tex] \sum_{x} \sum_{y} f(x,y) n^x m^y [/tex].
Now my question is how can I get the expectation of [tex] f(x,y) [/tex] from the above generating function?