|
Guest
|
f(x,y) is jointly convex in x and y.
x=min(a1+r1, a2+r2)
y=max(a1+r1, a2+r2)
where r1~N(0,1), r2~N(0,1) are independently distributed (Normal
distribution)
Then if we take the expectation of f(x,y) over r1, r2, is E(f(x,y))
quasi-convex in (a1, a2)?
Thanks! |
|
|