Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

QM-1a 2018

End term Solution (based on corrected Q1)

Q1. Writing Valid Balls as VB; P(VB) = 1- P(no ball)-P(wide ball) = 1- 0.04-0.02 = 0.94

a. P (out | VB) = =

= = 0.0532

b. P (six sixes in 6 consecutive valid ball) = P (6 consecutive balls are six and VB|6 consecutive
balls are VB)
= P (6 six) / P (6 VB)
= = 5.937 10-9
c. As balls are independent, so are valid balls. Therefore the required probability is P(four|VB) = P
(four) /P (VB) = 0.15/ 0.94 = 0.1594.

d. P (out|VB) = 0.0532 P (not out|VB) = 1-0.0532 = 0.9468

Write X = The number of outs in 10 valid balls ~ B (10, 0.0532)

P(More than 3 out in 10 VB | At least one out in those 10 VB) = P (X>3| X ≥1) = P(X≥4| X≥1)

P (X ≥ 4) = 1-P(X ≤ 3) = 1 –[ (0.9468)10 + 10C1 (0.9468)9 (0.0532) + 10


C2 (0.9468)8 (0.0532)2
10
+ C3 (0.9468)7 (0.0532)3]

=0.001297

P (X ≥ 1) = 1- (0.9468)10 = 0.42108

So, P (X ≥ 4 | X ≥ 1) = P (X ≥ 4)/P( X ≥ 1) = 0.001297 / 0.42108= 0.0031

e. Expected Runs in 1 ball


= 1 0.04 + 1 0.02 + 1 0.02 + 1 0.02 + 0 0.30 + 1 0.20 +2 0.10 + 3 0.05 + 4 0.15 + 5 0.01
+ 6 0.04 +0 0.05
= 1.54
Expected runs in 30 balls = 30 1.54 = 46.2

f. Expected Runs in 1 Valid ball


= (1 0.02 + 1 0.02 + 0 0.30 + 1 0.20 +2 0.10 + 3 0.05 + 4 0.15 + 5 0.01 + 6 0.04
+0 0.05)/0.94
= 1.574
Expected runs in 30 valid balls = 30 1.574 = 47.234
Q2.

Acceptable Ranges: C: 1.2% to 1.8%, S: 1.5% to 1.7%

Silicon\Carbon C<1.2 1.2<C<1.8 C>1.8 Marginal of S


S<1.5 0.02 0.05 0.01 0.08
1.5<S<1.7 0.30 0.40 0.10 0.80
S>1.7 0.02 0.05 0.05 0.12
Marginal of C 0.34 0.50 0.16 1

a. For independence we need P(S in Range A, C in Range B) =P(S in Range A) P(C in Range B)
for all possible cases.

However, observe that P(S<1.5, 1.2<C<1.8) = 0.05 ≠ 0.08 0.50 = P(S<1.5) P(1.2<c<1.8)

So, carbon content and silicon content are not independent.

b. Is = 1, 1.5%<S<1.7% IC = 1, 1.2%<C<1.8%
0, o.w 0, o.w.

IS\IC IC =0 IC =1 Marginal of IS
0 0.02+0.01+0.02+0.05 = 0.1 0.05+0.05 = 0.1 0.2
1 0.3+0.1 = 0.4 0.4 0.8
Marginal of IC 0.5 0.5 1

Clearly, P (Is =i, IC =j) =P (Is =i) P(IC =j) for i = 0, 1; j = 0,1.

Hence Is and IC are independent.

c. X= -1, C < 1.2% Y= -1, S<1.5%


0, 1.2<C<1.80, 0, 1.5<S<1.7%
1 C >1.8% 1 S>1.7%

Y\X -1 0 1 Marginal of Y
-1 0.02 0.05 0.01 0.08
0 0.30 0.40 0.10 0.80
1 0.02 0.05 0.05 0.12
Marginal of X 0.34 0.50 0.16 1

(i) E (XY) = (-1) (-1) 0.02+(1-) 1 0.01+1 (-1) 0.02+1 1 0.05 =0.07 – 0.03 = 0.04.
(ii) E (X2 + Y2) = E(X2) + E(Y2) = ((-1)2 0.34+12 0.16) + ((-1)2 0.08+12 0.12) = 0.5+0.2 =0.7.

(iii) X2 + Y2 = 1 when (X, Y) = (0,1), (1,0), (0,-1), (-1,0). This means that XY = 0 in all such cases.
Hence P(XY = 0| X2 + Y2 = 1) = 1.
Q.3. P (1) =2/6 = 1/3 P (2) = 3/6 = 1/2 P(3) = 1/6
a. If Xi = Value on the ith throw, then
(i) E (Xi) = 1 (2/6) + 2 (3/6) + 3 (1/6) =11/6
E (X1 + X2 +…+X100) = 100 E(X1) = 100 11/6 =183.33
(ii) V(X1) = E [(X1 –E(X1))2] = E(X12) – (E(X1))2
= 12 (2/6) + 22 (3/6) + 32 (1/6) – (11/6)2 = 23/6 – 121/36
= 17/36
V (X1 + X2 +…+X100) = 100V(X1) = 100 17/36 = 47.22
S.D. = √ = 6.87

(iii) By CLT, X1 + X2 +…+X100 N( ) approximately.

Hence, P (180 < X1 + X2 +…+X100 < 190) = ( )- ( )

= (0.97) - (-0.48) = 0.8340-0.3156 =0.5184.

(b) (X1, X2, X3) Multinomial (5, 2/6, 3/6, 1/6)

(i) Now, E (X2- X3) = E (X2) - E(X3).

But, X1 Bin (5, 2/6) X2 B (5, 3/6) X3 B (5,1/6)

E (X2) =5 3/6 =15/6; E (X3) = 5 1/6 = 5/6.

So, E (X2- X3) = 15/6 – 5/6 = 10/6 = 1.667

(ii) As X1+X2+X3 =5, X2+X3 ≤ 5.

Then,

For X2 ≥ 3, X2>X3 regardless the value of X3.

For X2 = 2, X2> X3 means either X3 = 1 or X3 = 0.

For X2 = 1, X2> X3 means X3 = 0.

For X2 = 0, X2> X3 is not possible.

Hence, P(X2>X3) = P(X2 ≥ 3) + P(X2 =2, X3 = 1 or 0) + P(X2 =1, X3 = 0)

= P(X2 ≥ 3) + P(X1 =2, X2 =2, X3 = 1) + P(X1 =3, X2 =2, X3 = 0) + P(X1 =3,X2 =1, X3 = 0)

 5  2   3   1    2   3 
2 2 3 2
5
=0.5 (by symmetry, P(X2 ≥ 3) = 0.5)+        +      +
 2 2 1  6   6   6   3 2 0   6   6 
 5  2   3 
4 1

    
 4 1 0 6   6 

= 0.5+ = 0.5 + 0.262 = 0.762.


Q 4:

Λ = 1/min

(a) Let X1, X2, X3, X4 be the number of customers arriving in the time intervals N[10:01—10:03],
N[10:04 —10:05], N[10: 06 —10:08] and N[10:09 —10:15] respectively.

Hence X1 ~ Poisson (3), X2 ~ Poisson (2), X3 ~ Poisson (3), X4 ~ Poisson (7) independently, and
hence

X1+X2+X3+X4 ~ Poisson (15)

Joint Distribution is given by the following pmf (for x1+x2+x3+x4 = 20)

P(X1 = x1, X2 = x2, X3 = x3, X4 = x4 | X1+X2+X3+X4 = 20)

e3 3x1 e 2 2 x 2 e 3 3x3 e 7 7 x 4


x1 ! x 2 ! x 3 ! x 4 ! 20! 3x1 2 x 2 3x3 7 x 4
= =
e 151520 x1 !x 2 !x 3 !x 4 ! 1520
20!

  3  1  2  2  3  3  7 
x x x x4
20
= 
 x1 x2 x3 x 4   15   15   15   15 

Which tells us that (X1, X2, X3, X4| X1+X2+X3+X4 = 20) is multinomial(20, 3/15, 2/15, 3/15, 7/15).

(b) P(N[10:01—10:07] = 5 and N[10:05 — 10:10] = 6 | N[10:01—10:10] = 10)


= P(N [10:01—10:04] = 4, N [10:05—10:07] = 1, [10:08—10:10] = 5)/P(N[10:01—10:10] = 10)

Now, N[10:01—10:04], N[10:05 —10:07] and N[10:08—10:10] are independent, and Poisson(4),
Poisson (3) and Poisson (3) respectively, While N[10:01—10:10] is Poisson(10).

Hence the required probability


e4 44 e3 31 e 3 35
10!
= 4! 101! 10 5!  0.44 0.310.35
e 10 4!1!5!
10!
=0.024.

(One can also directly bring in the multinomial distribution.)


Q 5. X~ U (-1, 1) Y~ U (-1, 1), independent.

a) Difference in length of two rods: X-Y. E(X-Y) = E(X)-E(Y) = 0-0 =0.

Note: absolute difference is given by |X-Y|. That is not what this problem asks for.

b) E[(X+Y)(X-2Y)] = E(X2+XY-2XY-2Y2) = E(X2) – E(XY) -2E(Y2)


=E(X2)-E(X)E(Y)-2E(Y2) (Using E(X) = E(Y) = 0, and independence of X and Y)
= V(X) – 2V(Y) (As E(X) = E(Y) = 0)
= - 1/3 (As V(X) = V(Y) = (1-(-1))2/12 = 4/12 = 1/3)

c) As X and Y are uniformly distributed on (-1,1), both are symmetric around 0. Hence so is X+Y.
This gives P(X+Y>0) = 0.5.
d) P(X>0 | X+Y>0) = P(X>0 ∩ X+Y > 0) / P(X+Y>0) = (3/8)/(1/2) = 0.75.
Justification: X+Y>0 is the blue region, and X>0 ∩
X+Y > 0 is the textured region. By symmetry, it is
obvious that P(X>0 ∩ X+Y > 0) = 3/8.)

Alternative argument: If both X and Y>0, then


X+Y> 0. When X<0 and Y<0, we can never have
X+Y>0. If X<0 and Y>0, i.e., in the 4th quadrant,
P(X+Y>0) = ½ by symmetry. The same is true in the
2nd quadrant, i.e., if X>0 and Y<0. Hence, P(X>0 ∩
X+Y > 0) = P(1st quadrant)+ ½ P(2nd quadrant) = 3/8.

Q 6. Time spent by each customer is Exponential(λ = 1/1 = 1)

Let the time spent by the 5 customers be X1, X2, X3, X4, X5. The Xi’s are independent Exponential(1).

P(Xi >2 hours) = S(2) = e-2 where S is the survival function of Exponential(1).
(a) P(all customers spend >2 hours) = (e-2)5 = e -10 by independence.

(b) P(Xi >2|Xi >1) = P(Xi>1) = e-1 by memoryless property of exponential.

Hence, P(Xi <2|Xi >1) = 1- P(Xi >2|Xi >1) = 1- e-1

So P (one customer spend time between 1hr and 2 hr and 4 spent more than 2 hours| all 5 spent more
than one hour) = 5C4 (e-1)4(1 – e-1) = 5(e-4-e-5)
Q 7. Left piece = T kg, Right piece = 10- T kg T ~ U (4, 6)

Case 1: If T is in (4, 5) then 10- T is in (5, 6), 10-T is then maximum, i.e., S = 10-T
Case 2: If T in in (5, 6) then 10-T is in (4, 5), T is then maximum, i.e., S = T.
Clearly, by symmetry, both cases have probability ½.
Also, as S is 10-T if T<5 and T if T>5, 5<S<6.

(a) The distribution of S is obtained from its distribution function. Now,


FS(s) = P(S≤s) = P(10-T≤s|T is in (4,5)) P(T is in (4,5)) + P(T≤s|T is in (5,6)) P(T is in (5,6))
(5<s<6)
= P(10-s≤T<5) ½ + P(5< T≤s) ½ = P(10-s≤T≤s) ½ (s-(10-s))/(6-4) = s-5 for 5<s<6.
Hence, the density of S is 1, in the range (5,6), and 0 otherwise. I.e., S~U(5,6).

Alternative approach: S= Max (T, 10-T).

FS(s) = P(S≤s) =P(Max(T, 10-T) ≤s) = P(Both T and (10-T) ≤s) = P( 10-s≤ T ≤s) = s-5 for 5<s<6.
Hence, S~U(5,6).

(b) E(S) = (5+6)/2 = 5.5


V(S) = (6-5)2 / 12 = 1/12 = 0.083

(c) Let X be the size of the piece Swapnil gets. Then X=S with probability 0.75, and 10-S with
probability 0.25.
Hence, E (X) = 0.75 E(S) + 0.25 E(10-S) = 0.25 10 + 0.5E(S) = 2.5 + 2.75 = 5.25 kg

You might also like