Student Information

Name : Emre Geçit

ID : 2521581

Answer 1
a) X and Y are independent ifff (x)f (y) = f (x, y).

This function has a nonzero value inside

√ √
− 1 − x2 < y < 1 − x2
p p
− 1 − y2 < x < 1 − y2.

Since integral of f (x, y) outside these boundaries is zero, integrating between these boundaries
will be enough.
R √
1−x2
f (y) = √ 1/π dx
− 1−x2
√
2 1−x2
f (y) = π
R√ 1−y 2
f (x) = √ 1/π dy
1−y 2
−
√
2 1−y 2
f (x) = π
√ √ √
2 1−y 2 2 1−x2 4 (1−x2 )(1−y 2 )
f (x)f (y) = π π
= π2

f (x)f (y) ̸= f (x, y)

Since f (x)f (y) is not equal to f (x, y), random variables X and Y are not independent.

√ √
2 1−x2 2 1−y 2
b) From a, marginal pdf for X is f (x) = π
and marginal pdf for Y is f (y) = π

R 1
R 1 √
2x 1−x2
c) Expected value for f (x) is E(X) = xf (x) dx = π
dx
−1 −1
−2(1−x2 )3/2 1
Solving this integral we find E(X) = 3π
|−1 =0

1
d) V ar(X) = E(X 2 ) − µ2
R1 R1 √
2x2 1−x2
E(X 2 ) = −1
x2 f (x) dx = −1 π
dx = 0.25

µ = E(X) = 0 (from part c)

V ar(X) = 0.25 − 02 = 0.25

2
Answer 2
a) Since these two events are independent f (tA , tB ) = f (tA )f (tB )

f (tA ) = 1/100 and f (tB ) = 1/100

f (tA , tB ) = 1/100 × 1/100 = 1/10000

R 10 R 100
b) P (tA < 10, tB > 90) = P (tA < 10)P (tB > 90) = 0
f (tA ) dt × 90
f (tB ) dt

t 10 t 100
= ( 100 |0 ) × ( 100 |90 ) = 0.1 × 0.1 = 0.01

c) P (tA < tB + 20)

This probability distribution can be shown with a 2d graph as follows:

100
tA

80

60

40

20
tB
20 40 60 80 100
The probability P (tA < tB + 20) is the ratio between the area under the curve tA < tB + 20 to
the probability space.

That is 6800/10000 = 0.68

3
d) Likewise the previous question, this probability can also be modelled as a 2d graph.

P (|tA − tB | < 30) :

|tA − tB | < 30 ≡ (tB − 30 < tA < tB + 30)

100
tA

80

60

40

20
tB
20 40 60 80 100
P (|tA − tB | < 30) = 5100/10000 = 0.51

Answer 3
a) Let FT be the cdf of the random variable T .

FT (t) = P (T < t)
= 1 − P (T ≥ t)
= 1 − P (min{X1 , X2 , ..., Xn } ≥ t)
= 1 − P (X1 ≥ t, X2 ≥ t, ..., XN ≥ t)
= 1 − P (X1 ≥ t)P (X2 ≥ t)...P (XN ≥ t)
= 1 − e−tλ1 −tλ2 −...−tλN
PN
= 1 − e− n=1 tλn

4
b) Let T be the first moment that any of the computers fails and Xn be the random variable
that represents the time that computer n fails.

T = min{X1 , X2 , ..., Xn }
PN
Using the formula from part a, we get FT (t) = 1 − e− n=1 tλn

λn = n/10
PN nt
We get, FT (t) = 1 − e− n=1 10 = 1 − e−55/10

This cdf in the form of a exponential cdf. Then λ = −(−55/10) = 55/10

E(T ) = 1/λ = 10/55 = 0.18

Answer 4
a) At least %70 of the participants are undergraduate students means that there are at least 70
undergraduate students among the participants.

The number U of
pundergraduate students has binomial distribution with n = 100 and p = 0.7,
µ = 74 and σ = 100 ∗ 0.74 ∗ (1 − 0.74) = 4.386.

Applying the central limit theorem with the continuity correction,

P (U ≥ 70) = P (U > 69.5) = P ( U σ−µ ≥ 69.5−µ

σ
)

69.5−74
P (Z > 4.386
)

P (Z > −1.03) = 1 − P (Z ≤ −1.03)

P (Z ≤ −1.03) = Φ(−1.03) = 0.1515

P (Z > −1.03) = 1 − P (Z ≤ −1.03) = 0.8485

The probability that at least %70 of participants are undergraduate students is 0.8485

5
b) At most %5 of the participants are pursuing a doctoral degree means that there is at most 5
doctoral students among the participants.

The number D of participants that areppursuing a doctoral degree has binomial distribution with
n = 100 and p = 0.1, µ = 10 and σ = 100 ∗ 0.1 ∗ (1 − 0.1) = 3

Applying the central limit theorem with the continuity correction,

P (D ≤ 5) = P (D < 5.5) = P ( D−µ

σ
≤ 5.5−µ
σ
)

5.5−10
P (Z ≤ 3
) = Φ(−1.5) = 0.0668

The probability that at most %5 of participants are pursuing a doctoral degree is 0.0668