Exam 1

STOR 612

Please only use the paper given by the instructor. In LP formulations, do not make
assumptions based on the problem data (ie. since product 1 is cheaper than product
2, we will not make any of product 1, etc.), even in the unlikely case that this would
be possible.

Please write your solutions clearly, and legibly.

Question 1 (20 points)

For an α real number, define the following functions:

f (α) = max { α, 2α − 3, 6α + 5 }
g(α) = max { α + 1, 3α + 5, 7α + 5 }
h(α) = min { α + 2, 3α + 7, 8α + 10 }.

We are given 2k row vectors a1 , . . . , ak ; b1 , . . . , bk . We are also given a matrix A and a

column vector d. From the following list of optimization problems, identify the ones
that can be formulated as an LP based om what we covered in class.

If a problem cannot be formulated, just say this. If a problem can be formulated,

say this, and give the formulation. Use the minimum number of variables up to a
constant factor (i.e. using 2k instead of k is OK; but using k 2 instead of k will not
receive full credit.) Make sure that all inequalities are in the correct direction. No
proof of correctness is necessary.

1. Pk
max f (ai x)
Pi=1
k
s.t. i=1 g(bi x) ≤ δ
Ax ≤ d.
Solution We cannot do this. We could represet the constraint, but not the
objective.

2. Pk
min f (ai x)
Pki=1
s.t. i=1 g(bi x) ≤ δ
Ax ≤ d.

1
 Use variables z1 , . . . , zk ; s1 , . . . , sk . The LP is
Pk
min i=1 zi
s.t. zi ≥ ai x, ∀i
zi ≥ 2ai x − 3, ∀i
zi ≥ 6ai x + 5 ∀i
P
i si ≤ δ
si ≥ bi x + 1 ∀i
si ≥ 3bi x + 5 ∀i
si ≥ 7bi x + 5
Ax ≤ d.

3. Pk
min f (ai x)
Pi=1
k
s.t. i=1 g(ai x) ≥ δ
Ax ≤ d.
Solution We cannot do this. We could represet the objective but not the con-
straint.

4. Pk
min f (ai x)
Pi=1
k
s.t. i=1 h(bi x) ≥ δ
Ax ≤ d.
Use the same zi and si as before, represent the objective as before, and the
constraint as
X
si ≤ bi x + 2, 3bi x + 7, 8bi x + 10 ∀i and si ≥ δ.
i

5. Pk
max h(ai x)
Pi=1
k
s.t. i=1 g(bi x) ≤ δ
Ax ≤ d.
Represent the constraint as before, and the objective as
X
max zi , zi ≤ 2bi x + 2, 3bi x + 7, 8bi x + 10 ∀i.
i

6. Pk
min h(ai x)
Pi=1
k
s.t. i=1 g(bi x) ≤ δ
Ax ≤ d.
We cannot do this.

2
Question 2 (20 points) Let R be a return matrix, with m rows, and n columns, with
components Rij , where Rij is the return on 1 dollar invested in stock j in scenario i.
Denote by pi (i = 1, . . . , m) the probability of scenario i happening. Let x be a
portfolio, i.e., x is a nonnegative n-vector, whose components sum to 1, with xj equal
to the amount invested in stock j.

1. Define the expected return of portfolio x, and the average downside risk.
2. Given a δ scalar, formulate the problem of maximizing the expected return subject
to the ADR being ≤ δ, as a linear program. What is the maximum expected
return that can be achieved, if we want the ADR to be zero?

Question 3 (15 points) A cargo plane has three compartments for storing cargo:
front, center, and back. These compartments have capacity limits, as follows: front, 12
tons; center, 18 tons; back, 10 tons.

The following three types of cargo have been offered for shipment on an upcoming
flight:

Cargo Max Weight (tons) Profit (dollar/ton)

1 20 320
2 16 400
3 25 360

Any portion of these cargoes can be accepted; i.e. from cargo 1 we can take any
amount between 0 and 20 tons. The objective is to determine how much (if any) of
each cargo should be accepted, and how to distribute each among the compartments
to maximize the total profit for the flight.

To maintain balance of the airplane, the weight loaded into the back must be be-
tween 45 and 55 % of the weight loaded into the center. Also, business considerations
dictate that at least 20 % of all accepted cargo by weight must be cargo 1; and the
total weight of accepted cargoes 1 and 2 must not exceed the weight of accepted cargo
3 by more than 20 %.

Formulate a linear programming problem to determine how many tons of each cargo
to ship to maximize profit (do not attempt to solve). Describe the meaning of each
constraint.

Solution Let us call the front, center, and back compartments 1, 2, and 3, respec-
tively.

• Variables: x11 is the amount (in tons) cargo 1 put into compartment 1; x12 is
the amount (in tons) of cargo 1 put into compartment 2; etc.
• Objective: Maximize 320x11 + 320x12 + 320x13 + 400x21 + 400x22 + 400x23
+360x31 + 360x32 + 360x33

3
 • Constraints: For the availability of the cargoes:

x11 + x12 + x13 ≤ 20;

x21 + x22 + x23 ≤ 16;
x31 + x32 + x33 ≤ 25;

For the compartments’s capacity:

x11 + x21 + x31 ≤ 12;

x12 + x22 + x32 ≤ 18;
x13 + x23 + x33 ≤ 10;

For the front-center balance:

x13 + x23 + x33 ≤ 0.55(x12 + x22 + x32 ); which is the same as

x13 + x23 + x33 − 0.55x12 − 0.55x22 − 0.55x32 ≤ 0
x13 + x23 + x33 ≥ 0.45(x12 + x22 + x32 ); which is the same as
x13 + x23 + x33 − 0.45x12 − 0.45x22 − 0.45x32 ≥ 0.

At least 20 percent of all accepted cargo is cargo 1:

x11 + x12 + x13 ≥ 0.2(x11 + x12 + x13 + x21 + x22 + x23 + x31 + x32 + x33 );

which is the same as

0.8x11 + 0.8x12 + 0.8x13 − 0.2x21 − 0.2x22 − 0.2x23 − 0.2x31 − 0.2x32 − 0.2x33 ≥ 0;

The total weight of accepted cargoes 1 and 2 must not exceed the weight of
accepted cargo 3 by more than 20 %:

x11 + x12 + x13 + x21 + x22 + x23 ≤ 1.2(x31 + x32 + x33 ),

which is the same as

x11 + x12 + x13 + x21 + x22 + x23 − 1.2x31 − 1.2x32 − 1.2x33 ≤ 0.

All variables nonnegative.

Question 4 (20 points total) Consider the following production process model below.
Please describe an LP to find the production plan that yields maximal profit.

• Raw material costs 4 $ per lb.

• At the cost of 3 $ 1 lb of raw material can be processed into 3 oz of product 1

AND 4 oz of product 2.

• From 1 oz of product 2, at the cost of 4 $, we can make 1/2 oz of product 3, OR

1/3 oz of product 4 (this means we can make quantities in between.)

4
• Product 2 and product 4 can be blended to make product 5. x oz of product 2
and y oz of product 4 will make x + y oz of product 5. Between 20 and 30 % of
product 5 must be of product 2.

• Any portion of any product can also be directly sold; 1 oz of product i sells for
pi dollars.

• At most 800 lb of raw material is available.