Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

MATH5165 OPTIMIZATION

ASSIGNMENT

Term 1, 2022

1. Consider the optimization problem

(P1)    min

xRn- s.t.

1

|x x x0 |2(2)  > 1

eT x = 1,

where n 之 2, x = (x1 , x2 , . . . , xn)T  | nn , x0 = ( , . . . , )T  | nn  and e = (1, 1, . . . , 1)T  | nn .

(i) Show that the problem (P1) is a convex optimization problem.

(ii) Using the Karush-Kuhn-Tucker optimality conditions, find the global minimizer of (P1).


2. Consider the following non-convex quadratic minimization problem

(P2 )

min

xRn-

s.t.

xT Ax + 2bT x + c

|x|2(2) x r > 0,

where A is a symmetric (n < n) constant matrix, b is a constant n < 1 vector, c is a scalar, |x|2 = ′xT x and r > 0 is a scalar. You are given that x*  | nn , λ*  | n and the following conditions hold:

(A + λ I)x** + b = 0,    |x* |2(2) x r > 0,   λ* (|x* |2(2) x r) = 0,    (A + λ* I) e 0,   λ*  之 0,

where I is the (n < n) identity matrix and (A + λ* I) e 0 means that the matrix (A + λ* I) is a positive semi-definite matrix. Note that the matrix A is not assumed to be positive semi-definite. Prove that x* is a global minimizer of (P2).

Hint: Show first that the Lagrangian function L(x, λ* ) of (P2) is a convex function on nn .


3. Minimum Cost Pizza Problem. Using only the items given in the tables below, formulate an optimization problem in standard form to create a minimum cost pizza which satisfies both the nutritional requirements of Table 1 and bounds on item quantities given in Table 2. Use the nutritional data of Table

3 and the cost data of Table 4 in your model.

Use the MATLAB linear optimization routines linprog to solve the problem. Interpret your results.

Table 1

Nutrient

Requirement

Units

Calcium

750.0

mg

Iron

12.0

mg

Protein

48.5

gram

Vitamin A

4500.0

IU

Thiamine

1.3

mg

Niacin

16.0

mg

Riboavin

1.6

mg

Vitamin C

30.0

mg

Table 2

UPPER AND LOWER BOUNDS ON PIZZA ITEMS

Item

Upper Bounds*

Lower Bound*

Sauce

1.986

1.140

dough

5.249

4.266

cheese

2.270

1.703

pepperoni

0.983

N/A

ham

1.135

N/A

bacon

0.993

N/A

g.pepper

1.561

N/A

onion

0.993

N/A

celery

1.561

N/A

mushroom

1.135

N/A

tomato

1.703

N/A

pineapple

1.703

N/A

meat

N/A

0.993

veg.

N/A

0.993

fungi

N/A

0.922

* Amount in hundreds of grams.










4. Portfolio Selection Problem. An individual with ✩100,000 to invest has identified three mutual funds as attractive opportunities. Over the last five years, dividend payments (in cents per dollar invested) have been as shown in Table 5, and the individual assumes that these payments are indicative of what can be expected in the future. This particular individual has two requirements:   (1)  the combined expected yearly return from his/her investments must be no less than ✩8,000 (the amount ✩100,000 would earn at 8 percent interest) and   (2)  the variance in future, yearly, dividend payments should be as small as possible.  How much should this individual fully invest his/her ✩100,000 in each fund to achieve these requirements?

Table 5

Years

1

2

3

4

5

Investment 1

10

4

12

13

6

Investment 2

6

9

6

5

9

Investment 3

17

1

11

19

2

[Hint: Let xi(i = 1, 2, 3) designate the amount of funds to be allocated to investment i, and let xik denote the return per dollar invested from investment i during the kth time period in the past (k = 1, 2, . . . , 5). If the past history of payments is indicative of future performance, the expected return per dollar from

investment i is

5

Ei = xik .

k=1

The variance in future payments can be expressed as

3      3

f (x1 , x2 , x3 ) =             σij(2)xixj  = xT Cx,

i=1 j=1

where the covariances σij(2)  are given by

σij(2)  = k 1 xikxjk x ╱k 1 xik\ ╱k 1 xjk\ .   ]

(a) Using the following table, calculate the covariance matrix C = [σij(2)].