Unformatted text preview: hen X and Y
are independent. In the discrete case this is easy:
P (X + Y = z ) =
P (X = x)P (Y = z x)
x To see the ﬁrst equality, note that if the sum is z then X must take on some
value x and Y must be z x. The ﬁrst equality is valid for any random variables.
The second holds since we have supposed X and Y are independent. 215 A.3. EXPECTED VALUE, MOMENTS Example A.16. If X = binomial(n, p) and Y = binomial(m, p) are independent, then X + Y = binomial(n + m, p).
Proof by direct computation.
P (X + Y = i) = i
X ✓n ◆
j =0 j = p (1 p) i = ✓ p (1
j n+m i ◆
i p) nj · ✓ m
i j ◆ pi j (1 p)m i+j i
X ✓n ◆ ✓ m ◆
j =0 p)n+m i The last equality follows from the fact that if we pick i individuals from a group
of n boys and m girls, which can be done in n+m ways, then we must have j
boys and i j girls for some j with 0 j i. Much easier proof. Consider a sequence of n + m independent trials. Let
X be the number of successes in the ﬁrst n trials an...
View Full Document
This document was uploaded on 03/06/2014 for the course MATH 4740 at Cornell.
- Spring '10
- The Land