Professional Documents
Culture Documents
TD2DS
TD2DS
INSAT
Department of Mathematics & Computer Sciences
Exercise 1: Let (X, Y ) be a random pair and below is its joint probability distribution
X⧹Y -1 0 1
1 1/8 1/4 1/8
2 1/4 1/8 1/8
X +Y −1 0 1
1 0 1 2
2 1 2 3
ans so
n o n o
X + Y ≤ 1 = (X, Y ) = (1, −1), (X, Y ) = (1, 0), (X, Y ) = (2, −1) ,
then
n o n o
P X +Y ≤1 = P (X, Y ) = (1, −1), (X, Y ) = (1, 0), (X, Y ) = (2, −1)
n o n o n o
= P (X, Y ) = (1, −1) + P , (X, Y ) = (1, 0) + P (X, Y ) = (2, −1)
= 1/8 + 1/4 + 1/4.
3. Determine the mass distribution function of Z = X + Y . You just need to ll up the next
table
Z =X +Y PZ n o n o
0 P{Z = 0} = P X + Y = 0 = P (X, Y ) = (1, −1) = 1/8
n o n o
1 P{Z = 1} = P X + Y = 1 = P (X, Y ) = (1, 0) or (X, Y ) = (2, −1) = 1/4 + 1/4
2 ···
3 ···
Finally we get ρXY = −0, 1601 and ρ2XY = 0, 0256. We note that ρ2XY ≪ 1 which means
that there is actually no linear correlation between X and Y or, in dierent terms,
X couldn't explain Y linearly.
Note:
Since we now know that E(Y |X) = φ(X) is linear we can use the values table of φ above, to
determine its expression. If we denote φ(x) = βx + α, we get
φ(1) = β + α = 0 β = −1/4
=⇒ .
φ(2) = 2β + α = −1/4 α = 1/4
Exercise 3:
First of all we should note that X and Y are correlated by construction since
X + Y ≤ 4.
1. Determine the joint mass distribution function of the random pair (X, Y ): for m + n ≤ 4
P{X = m, Y = n} = C4m C4−m
n
(90%)m (8%)n (2%)4−(m+n)
4! (4 − m)!
= (90%)m (8%)n (2%)4−(m+n)
m!(4 − m)! n!(4 − (m + n))!
4!
= (90%)m (8%)n (2%)4−(m+n)
m!n!(4 − (m + n))!
Finally we obtain the joint probability distribution
Y /X 0 1 2 3 4
0 0.0000 0.0000 0.0000 0.0000 0.0000
1 0.0000 0.0000 0.0000 0.0000 0.0000
2 0.0000 0.0001 0.0000 0.0000 0.0000
3 0.0000 0.0012 0.0097 0.0000 0.0000
4 0.0001 0.0023 0.0389 0.2916 0.0000
V E(Y |X)
RY2 |X =
V(Y )
V 80%(4 − X)
=
8%92%4
(80%)2 V(X) (80%)2 90%10%4 (80%)2 90%10%
= = =
8%92%4 8%92%4 8%92%
≃ 0, 0736 = 7.36% ≪ 1.
This means that although, the dependence is linear between X and Y , but it is very weak
and/or insignicant.
5. Compute the correlation coecient of X and Y , ρXY . Comment.
Actually we don't need to compute the correlation coecient, since
√ E(Y |X) is linear in
X , which means that ρ2XY = RY2 |X = 0.782%, and then ρXY = − 0.782% = −8.84%.
1 If you want more details about the proof you can ask your colleague!
Actually, there exists a closed form for P{N = n} which is out of the scope of this course.
4. Determine the random variable E(N |T ).
Since (N |T = k) ∼ B(k; p) its expectations expresses E(N |T = k) = kp and so E(N |T ) =
pT .