Chapter - 10 - (Un) Constrained - Optimization - LMS - Answers

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 80

CHAPTER 10

(UN)CONSTRAINED
OPTIMIZATION

Alexandra Kis

December 2020
OUTLINE
This chapter

1. Unconstrained optimization of a
function of two variables
2. Constrained optimization
i. The substitution method
ii. The Lagrange’s method
TIMELINE

Last class Today Next class

Chapter 9 Chapter 10 Chapter 19


▪ Partial derivatives ▪ Unconstrained optimization ▪ Integration
▪ Constrained optimization
▪ The substitution method
▪ The Lagrange’s method
OBJECTIVES FOR THIS CHAPTER

At the end of this chapter, you should be able to:


▪ Use the first-order partial derivatives to find the
critical points of a function of two variables.
▪ Use the second-order derivatives to classify the
critical points of a function of two variables.
▪ Use the substitution method to solve constrained
optimization problems.
▪ Use the Lagrange’s method to solve constrained
optimization problems.
Unconstrained
optimization
UNCONSTRAINED OPTIMIZATION

Minimum point Maximum point


𝑑𝑓 𝑥 𝑑2 𝑓 𝑥 𝑑𝑓 𝑥 𝑑2 𝑓 𝑥
= 0 and >0 = 0 and <0
𝑑𝑥 𝑑𝑥 2 𝑑𝑥 𝑑𝑥 2

Slope = 0 and convex Slope = 0 and concave

𝑦 𝑦

𝑓(𝑥)
𝑓(𝑥)

𝑥 𝑥
UNCONSTRAINED OPTIMIZATION

With
Minimum point onlyMaximum
one point
𝑑𝑓 𝑥
𝑑𝑥
= 0 and
𝑑2 𝑓 𝑥
𝑑𝑥 2
variable!!
>0
𝑑𝑓 𝑥
𝑑𝑥
= 0 and
𝑑2 𝑓 𝑥
<0
𝑑𝑥 2

Slope = 0 and convex Slope = 0 and concave

𝑦 𝑦
What
𝑓(𝑥)
about a
function with 2 𝑓(𝑥)

variables?
𝑥 𝑥
UNCONSTRAINED OPTIMIZATION
UNCONSTRAINED OPTIMIZATION

𝑧
𝑦 𝑧
𝜕𝑧
=0
𝑥 𝑑𝑥
𝑥
UNCONSTRAINED OPTIMIZATION

𝑧
𝑦 𝑧
𝜕𝑧
=0
𝑥 𝑑𝑥
𝑥

𝑧
𝜕𝑧
=0
𝑑𝑦
𝑦
UNCONSTRAINED OPTIMIZATION

CRITICAL POINT(S)

With 1 variable With 2 variables


𝑑𝑓 𝑥 𝜕𝑓 𝑥,𝑦 𝜕𝑓 𝑥,𝑦
=0 = 0 and =0
𝑑𝑥 𝜕𝑥 𝜕𝑦
Slope is flat The plan is flat

𝑦 𝑓(𝑥)

𝑥
UNCONSTRAINED OPTIMIZATION

How do you translate the convex/concave condition for a


minimum/maximum point to 2 variables?
UNCONSTRAINED OPTIMIZATION

𝜕2𝑓 𝜕2𝑓 𝜕2𝑓 𝜕2𝑓


= 𝑓𝑥𝑥 = 𝑓𝑦𝑦 = 𝑓𝑥𝑦 = 𝑓𝑦𝑥
𝜕𝑥 2 𝜕𝑦 2 𝜕𝑥𝜕𝑦 𝜕𝑦𝜕𝑥

Hessian matrix: 𝐻=
UNCONSTRAINED OPTIMIZATION

We need to find the determinant 𝜀 of this matrix…

2
𝜕 𝑓 𝜕2𝑓
= 𝑓𝑥𝑥 = 𝑓𝑥𝑦
𝜕𝑥 2 𝜕𝑥𝜕𝑦
𝐻=
𝜕2𝑓 𝜕2𝑓
= 𝑓𝑦𝑥 2
= 𝑓𝑦𝑦
𝜕𝑦𝜕𝑥 𝜕𝑦

𝜀 = det 𝐻 =
Note: 𝜀 is a Greek letter called “epsilon”
UNCONSTRAINED OPTIMIZATION

We need to find the determinant 𝜀 of this matrix…

2
𝜕 𝑓 𝜕2𝑓
= 𝑓𝑥𝑥 = 𝑓𝑥𝑦
𝜕𝑥 2 𝜕𝑥𝜕𝑦
𝐻=
𝜕2𝑓 𝜕2𝑓
= 𝑓𝑦𝑥 2
= 𝑓𝑦𝑦
𝜕𝑦𝜕𝑥 𝜕𝑦

𝜀 = det 𝐻 = 𝑓𝑥𝑥 𝑓𝑦𝑦


UNCONSTRAINED OPTIMIZATION

We need to find the determinant 𝜀 of this matrix…

2
𝜕 𝑓 𝜕2𝑓
= 𝑓𝑥𝑥 = 𝑓𝑥𝑦
𝜕𝑥 2 𝜕𝑥𝜕𝑦
𝐻=
𝜕2𝑓 𝜕2𝑓
= 𝑓𝑦𝑥 2
= 𝑓𝑦𝑦
𝜕𝑦𝜕𝑥 𝜕𝑦

𝑓𝑥𝑦 = 𝑓𝑦𝑥
2
𝑓 𝑓 −𝑓
𝜀 = det 𝐻 = 𝑥𝑥 𝑦𝑦 𝑥𝑦
UNCONSTRAINED OPTIMIZATION

CRITICAL POINT(S) +

With 2 variables
With 1 variable
𝑑2 𝑓 𝑥 𝑓𝑥𝑥 > 0 → minimum
> 0 → minimum
𝑑𝑥 2
𝜀>0
𝑑2 𝑓 𝑥
< 0 → maximum 𝑓𝑥𝑥 < 0 → maximum
𝑑𝑥 2
𝑑2 𝑓 𝑥
= 0 → inflexion 𝜀 < 0 → saddle point
𝑑𝑥 2

𝜀 = 0 → no information
UNCONSTRAINED OPTIMIZATION
UNCONSTRAINED OPTIMIZATION

Hessian matrix
Matrix of the second-order partial derivatives.
With 𝑓(𝑥, 𝑦):
𝜕2𝑓 𝜕2𝑓
𝜕𝑥 2 𝜕𝑥𝜕𝑦 𝑓𝑥𝑥 𝑓𝑥𝑦
𝐻= =
𝜕2𝑓 𝜕2𝑓 𝑓𝑦𝑥 𝑓𝑦𝑦
𝜕𝑦𝜕𝑥 𝜕𝑦 2
UNCONSTRAINED OPTIMIZATION

Determinant 𝜀 of the Hessian matrix


With:
𝑓𝑥𝑥 𝑓𝑥𝑦
𝐻=
𝑓𝑦𝑥 𝑓𝑦𝑦

2
𝜀 = det 𝐻 = 𝑓𝑥𝑥 𝑓𝑦𝑦 − 𝑓𝑥𝑦
UNCONSTRAINED OPTIMIZATION

Optimization of a function of two variables


With a function 𝑓(𝑥, 𝑦):
1. Find the critical point by solving the system:
𝜕𝑓 𝜕𝑓
=0 and =0
𝜕𝑥 𝜕𝑦

2. Determine the nature of the critical point by computing the


2
determinant 𝜀 of the Hessian matrix → 𝜀 = 𝑓𝑥𝑥 𝑓𝑦𝑦 − 𝑓𝑥𝑦
3. If 𝜀 > 0 and 𝑓𝑥𝑥 < 0 → maximum point
If 𝜀 > 0 and 𝑓𝑥𝑥 > 0 → minimum point
If 𝜀 < 0 → saddle point (neither a max, nor a min)
If 𝜀 = 0 → no information
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

1. Solve the system: 𝑓𝑥 = 0 and 𝑓𝑦 = 0

𝑓𝑥 = 3𝑥 2 − 3 + 𝑦 2 = 0

𝑓𝑦 = 2𝑥𝑦 = 0 Either 𝑥 = 0 or 𝑦 = 0
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

1. Solve the system: 𝑓𝑥 = 0 and 𝑓𝑦 = 0

𝑓𝑥 = 3𝑥 2 − 3 + 𝑦 2 = 0

𝑓𝑦 = 2𝑥𝑦 = 0 Either 𝑥 = 0 or 𝑦 = 0

If 𝑥 = 0 → 𝑦2 = 3 → 𝑦=± 3 (0, − 3) and (0, 3)


UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

1. Solve the system: 𝑓𝑥 = 0 and 𝑓𝑦 = 0

𝑓𝑥 = 3𝑥 2 − 3 + 𝑦 2 = 0

𝑓𝑦 = 2𝑥𝑦 = 0 Either 𝑥 = 0 or 𝑦 = 0

If 𝑥 = 0 → 𝑦2 = 3 → 𝑦=± 3 (0, − 3) and (0, 3)

If 𝑦 = 0 → 3(𝑥 2 − 1) = 0 → 𝑥=± 1 (−1, 0) and (1, 0)


UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

2. Find the Hessian matrix and its determinant 𝜀

𝑓𝑥 = 3𝑥 2 − 3 + 𝑦 2

𝑓𝑦 = 2𝑥𝑦
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

2. Find the Hessian matrix and its determinant 𝜀

𝑓𝑥𝑥 = 6𝑥
𝑓𝑥 = 3𝑥 2 −3 + 𝑦2

𝑓𝑦 = 2𝑥𝑦
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

2. Find the Hessian matrix and its determinant 𝜀

𝑓𝑥𝑥 = 6𝑥
𝑓𝑥 = 3𝑥 2 −3 + 𝑦2
𝑓𝑦𝑦 = 2𝑥
𝑓𝑦 = 2𝑥𝑦
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

2. Find the Hessian matrix and its determinant 𝜀

𝑓𝑥𝑥 = 6𝑥
𝑓𝑥 = 3𝑥 2 −3 + 𝑦2
𝑓𝑦𝑦 = 2𝑥
𝑓𝑦 = 2𝑥𝑦
𝑓𝑥𝑦 = 𝑓𝑦𝑥 = 2𝑦
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

2. Find the Hessian matrix and its determinant 𝜀

𝑓𝑥𝑥 = 6𝑥
𝑓𝑥 = 3𝑥 2 −3 + 𝑦2
𝑓𝑦𝑦 = 2𝑥
𝑓𝑦 = 2𝑥𝑦
𝑓𝑥𝑦 = 𝑓𝑦𝑥 = 2𝑦

6𝑥 2𝑦
𝐻= 𝜀 = det 𝐻 = 6𝑥 ∗ 2𝑥 − 2𝑦 ∗ 2𝑦 = 12𝑥 2 − 4𝑦 2
2𝑦 2𝑥
UNCONSTRAINED OPTIMIZATION

Example 1:
Find and classify the stationary point(s) of the following function:
𝑓 𝑥, 𝑦 = 𝑥 3 − 3𝑥 + 𝑥𝑦 2

3. Nature of the critical point with 𝜀 = 12𝑥 2 − 4𝑦 2 and 𝑓𝑥𝑥 = 6𝑥

(0, − 3) 𝜀 = −12 𝜀 < 0 → saddle point


(0, 3) 𝜀 = −12 𝜀 < 0 → saddle point
(−1, 0) 𝜀 = 12 and 𝑓𝑥𝑥 = −6 𝜀 > 0, 𝑓𝑥𝑥 < 0 → maximum
(1, 0) 𝜀 = 12 and 𝑓𝑥𝑥 = 6 𝜀 > 0, 𝑓𝑥𝑥 > 0 → minimum
Constrained
optimization
CONSTRAINED OPTIMIZATION
CONSTRAINED OPTIMIZATION

Maximize your consumption… Under your budget constraint

Maximize your profit… Under the constraint to produce a certain quantity

Minimize your cost… Under the constraint to produce a certain quantity

Maximize the quantity produced… Under the cost constraint


CONSTRAINED OPTIMIZATION

2 methods

The substitution method The Lagrange’s method

Only for 2 variables and 1 constraint…


The substitution
method
THE SUBSTITUTION METHOD

The substitution method


1. From the constrained equation, find an expression for one of the
unknown variables.
2. Replace this expression in the function to optimize.
3. Find the first-order derivative of the function to optimize and set it
equal to zero.
4. Solve for the unknown.
5. Check that it is a max/min with the second-order derivative.
6. Replace the value you found in step 4 into the expression you found
in step 1 and solve for the second unknown.
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2 Function to optimize
Subject to 2𝑄1 + 𝑄2 = 100 Constraint

1. From the constrained equation, find an


expression for one of the unknown variables.

𝑄2 = 100 − 2𝑄1
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100

2. Replace this expression in the function to optimize.

𝐶 = 𝑄1 𝑄2 = 𝑄1 (100 − 2𝑄1 )

𝐶 = 100𝑄1 − 2𝑄12 Depends only on 𝑄1 !


THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100

3. Find the first-order derivative of the function


to optimize and set it equal to zero.

𝑑𝐶
𝐶= 100𝑄1 − 2𝑄12 = 100 − 4𝑄1 = 0
𝑑𝑄1
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100

4. Solve for the unknown.

𝑑𝐶
= 100 − 4𝑄1 = 0 𝑄1 = 25
𝑑𝑄1
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100

5. Check it is a maximum.

𝑑𝐶
= 100 − 4𝑄1 = 0 𝑄1 = 25
𝑑𝑄1
𝑑2 𝐶
2 = −4 < 0 → 𝑚𝑎𝑥
𝑑𝑄1
THE SUBSTITUTION METHOD

Example 2:
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100

6. Replace the value you found in step 4 into the expression


you found in step 1 and solve for the second unknown.

𝑄2 = 100 − 2𝑄1 𝑄1 = 25 𝑄2 = 50
The Lagrange’s
method
THE LAGRANGE’S METHOD

The Lagrange’s method


With 𝑓(𝑥, 𝑦) the function to optimize and 𝑔 𝑥, 𝑦 = 𝑐 the
constraint, the Lagrange function is:
𝐿 𝑥, 𝑦 = 𝑓 𝑥, 𝑦 − λ ∙ (𝑔 𝑥, 𝑦 − 𝑐)
Then solve the system:
𝜕𝐿 𝜕𝐿 𝜕𝐿
= 0, =0 and =0
𝜕𝑥 𝜕𝑦 𝜕λ

Note: λ is a Greek letter called “lambda”


THE LAGRANGE’S METHOD

Example 2 (bis):
You want to maximize your consumption under your budget constraint:
max 𝐶 𝑄1 , 𝑄2 = 𝑄1 𝑄2
Subject to 2𝑄1 + 𝑄2 = 100 2𝑄1 + 𝑄2 − 100 = 0

The Lagrange function is: 𝐿 𝑄1 , 𝑄2 = 𝑄1 𝑄2 − λ (2𝑄1 + 𝑄2 − 100)


𝜕𝐿
= 𝑄2 − 2λ = 0 𝑄2 − 2𝑄1 = 0 𝑄2 = 2𝑄1
𝜕𝑄1
𝜕𝐿
= 𝑄1 − λ = 0 λ = 𝑄1
𝜕𝑄2
𝜕𝐿
= − (2𝑄1 + 𝑄2 − 100) = 0 100 − 2𝑄1 − 2𝑄1 = 0 𝑄1 = 25 𝑄2 = 50
𝜕λ
THE LAGRANGE’S METHOD

Do we have to check the nature of the critical point with


the Lagrange’s method?
Exercises
EXERCISES

Exercise 1:
A firm sells two goods 𝐺1 and 𝐺2 at $1000
and $800, respectively. The total cost of
producing these two goods is given by
𝐶 = 2𝑄12 + 2𝑄1 𝑄2 + 𝑄22
Where 𝑄1 and 𝑄2 denotes the output
levels of 𝐺1 and 𝐺2 , respectively. Find the
maximum profit and the values of 𝑄1 and
𝑄2 at which this is achieved.
EXERCISES

Exercise 1:
A firm sells two goods 𝐺1 and 𝐺2 at $1000 and $800, respectively. The total
cost of producing these two goods is given by
𝐶 = 2𝑄12 + 2𝑄1 𝑄2 + 𝑄22
Where 𝑄1 and 𝑄2 denotes the output levels of 𝐺1 and 𝐺2 , respectively. Find the
maximum profit and the values of 𝑄1 and 𝑄2 at which this is achieved.

𝑅 = 𝑃1 𝑄1 + 𝑃2 𝑄2 = 1000𝑄1 + 800𝑄2

𝜋 = 𝑅 − 𝐶 = 1000𝑄1 + 800𝑄2 − (2𝑄12 + 2𝑄1 𝑄2 + 𝑄22 )

𝜋 = 1000𝑄1 + 800𝑄2 − 2𝑄12 − 2𝑄1 𝑄2 − 𝑄22


EXERCISES

Exercise 1: 𝜋 = 1000𝑄1 + 800𝑄2 − 2𝑄12 − 2𝑄1 𝑄2 − 𝑄22


1. Critical points

𝜕𝜋
= 1000 − 4𝑄1 − 2𝑄2 = 0
𝜕𝑄1
200 − 2𝑄1 = 0 𝑄1 = 100
𝜕𝜋
= 800 − 2𝑄1 − 2𝑄2 = 0
𝜕𝑄2
𝑄2 = 300
EXERCISES

Exercise 1: 𝜋 = 1000𝑄1 + 800𝑄2 − 2𝑄12 − 2𝑄1 𝑄2 − 𝑄22


2. Hessian matrix

𝜋11 = −4
−4 −2
𝜋22 = −2 𝐻= 𝜀=4
−2 −2
𝜋12 = 𝜋21 = −2

3. Nature of the critical point

𝑄1 = 100
and 𝜀 = 4 and 𝜋11 = −4 𝜀 > 0, 𝜋11 < 0 → maximum
𝑄2 = 300
EXERCISES

Exercise 1: 𝜋 = 1000𝑄1 + 800𝑄2 − 2𝑄12 − 2𝑄1 𝑄2 − 𝑄22

𝜋 100, 300 = 1000 ∗ 100 + 800 ∗ 300 − 2 ∗ 1002 − 2 ∗ 100 ∗ 300 − 3002

= $170 000
EXERCISES

Exercise 2:
A firm produced two goods A and B. The
weekly cost of producing 𝑥 items of A and 𝑦
items of B is
𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
a) State the minimum value of cost in the
case when there are no constraints.
b) Find the minimum value of cost when the
firm is committed to producing 500 goods
of either type in total.
EXERCISES

Exercise 2:
A firm produced two goods A and B. The weekly cost of producing 𝑥 items of A
and 𝑦 items of B is
𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
a) State the minimum value of cost in the case when there are no constraints.

The higher you produce, the higher is the cost


→ the minimum cost if for 𝑥 = 0 and 𝑦 = 0
→ the minimum cost is 𝐶 = 1000
EXERCISES

Exercise 2:
A firm produced two goods A and B. The weekly cost of producing 𝑥 items of A
and 𝑦 items of B is
𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
b) Find the minimum value of cost when the firm is committed to producing
500 goods of either type in total.
min 𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
Constraint: 𝑥 + 𝑦 = 500 → 𝑥 = 500 − 𝑦
𝐶 = 0.2 500 − 𝑦 2 + 0.05𝑦 2 + 0.1(500 − 𝑦)𝑦 + 2(500 − 𝑦) + 5𝑦 + 1000

𝐶 = 52 000 − 147𝑦 + 0.15𝑦 2


EXERCISES

Exercise 2:
A firm produced two goods A and B. The weekly cost of producing 𝑥 items of A
and 𝑦 items of B is
𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
b) Find the minimum value of cost when the firm is committed to producing
500 goods of either type in total.

𝐶 = 52 000 − 147𝑦 + 0.15𝑦 2


𝑑𝐶
= −147 + 0.3𝑦 = 0 𝑦 = 490
𝑑𝑦
𝑥 = 500 − 𝑦 𝑥 = 10
EXERCISES

Exercise 2:
A firm produced two goods A and B. The weekly cost of producing 𝑥 items of A
and 𝑦 items of B is
𝐶 = 0.2𝑥 2 + 0.05𝑦 2 + 0.1𝑥𝑦 + 2𝑥 + 5𝑦 + 1000
b) Find the minimum value of cost when the firm is committed to producing
500 goods of either type in total.
𝑑𝐶 𝑑2 𝐶
= −147 + 0.3𝑦 2
= 0.3 > 0 → 𝑚𝑖𝑛
𝑑𝑦 𝑑𝑦

When 𝑥 = 10 and 𝑦 = 490:


𝐶 = 0.2 ∗ 102 + 0.05 ∗ 4902 + 0.1 ∗ 10 ∗ 490 + 2 ∗ 10 + 5 ∗ 490 + 1000
𝐶 = 15 985
EXERCISES

Exercise 3:
A monopolistic producer of two goods, 𝐺1 and 𝐺2 , has a total cost function
𝑇𝐶 = 5𝑄1 + 10𝑄2
Where 𝑄1 and 𝑄2 denote the quantities of 𝐺1 and 𝐺2 , respectively. If 𝑃1 and 𝑃2
denote the corresponding prices, then the demand equations are:
𝑃1 = 50 − 𝑄1 − 𝑄2
𝑃2 = 100 − 𝑄1 − 4𝑄2
a) Use the Lagrange method to find the maximum profit if the firm’s total costs
are fixed at $100.
b) Estimate the new optimal profit if the total costs rise to $101.
EXERCISES

Exercise 3:
𝑃1 = 50 − 𝑄1 − 𝑄2
𝑃2 = 100 − 𝑄1 − 4𝑄2
a) Find the maximum profit if the firm’s total costs are fixed at $100.

𝑅 = 𝑄1 𝑃1 + 𝑄2 𝑃2 𝑅 = 𝑄1 50 − 𝑄1 − 𝑄2 + 𝑄2 (100 − 𝑄1 − 4𝑄2 )

𝑅 = 50𝑄1 − 𝑄12 − 𝑄1 𝑄2 + 100𝑄2 − 𝑄1 𝑄2 − 4𝑄22

𝑅 = 50𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 100𝑄2 − 4𝑄22

𝜋 = 𝑅 − 𝐶 = 50𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 100𝑄2 − 4𝑄22 − 5𝑄1 − 10𝑄2


EXERCISES

Exercise 3:
𝑃1 = 50 − 𝑄1 − 𝑄2
𝑃2 = 100 − 𝑄1 − 4𝑄2
a) Find the maximum profit if the firm’s total costs are fixed at $100.

max 𝜋 = 45𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 90𝑄2 − 4𝑄22


s.t. 5𝑄1 + 10𝑄2 = 100

𝐿 = 45𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 90𝑄2 − 4𝑄22 − λ(5𝑄1 + 10𝑄2 − 100)


EXERCISES

Exercise 3:
a) Find the maximum profit if the firm’s total costs are fixed at $100.

𝐿 = 45𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 90𝑄2 − 4𝑄22 − λ(5𝑄1 + 10𝑄2 − 100)

𝜕𝐿 90 − 4𝑄1 − 4𝑄2 − 10λ = 0


= 45 − 2𝑄1 − 2𝑄2 − 5λ = 0
𝜕𝑄1
−2𝑄1 + 90 − 8𝑄2 − 10λ = 0
𝜕𝐿
= −2𝑄1 + 90 − 8𝑄2 − 10λ = 0 −2𝑄1 + 4𝑄2 = 0
𝜕𝑄2
𝜕𝐿 𝑄1 = 2𝑄2
= −(5𝑄1 + 10𝑄2 − 100) = 0
𝜕λ

−10𝑄2 − 10𝑄2 + 100 = 0 𝑄2 = 5 𝑄1 = 10 λ=3


EXERCISES

Exercise 3:
a) Find the maximum profit if the firm’s total costs are fixed at $100.

𝑄1 = 10, 𝑄2 = 5 𝜋 = 50𝑄1 − 𝑄12 − 2𝑄1 𝑄2 + 100𝑄2 − 4𝑄22 − 5𝑄1 − 10𝑄2

𝜋 = 600
EXERCISES

Exercise 3:
b) Estimate the new optimal profit if the total costs rise to $101.

How evolves the profit when the constraint increases (by 1)?

𝐿 = 𝑓(𝑥, 𝑦) − λ(𝑔 𝑥, 𝑦 − 𝑐)

𝜕𝐿
=λ‼
𝜕𝑐

λ=3

If the constraint (i.e. the total cost) increases by 1, the profit increases by 3.
EXERCISES

Exercise 4:
A consumer’s utility function is
𝑈 = ln 𝑥1 + 2ln(𝑥2 )
One good 𝑥1 costs 2 CHF and
one good 𝑥2 costs 3 CHF. You
can spend a maximum of 18 CHF
on both goods. Find the values of
𝑥1 and 𝑥2 which maximize 𝑈
subject to the budgetary
constraint.
EXERCISES

Exercise 4:
A consumer’s utility function is
𝑈 = ln 𝑥1 + 2ln(𝑥2 )
One good 𝑥1 costs 2 CHF and one good 𝑥2 cost 3 CHF. You can spend a
maximum of 18 CHF on both goods. Find the values of 𝑥1 and 𝑥2 which
maximize U subject to the budgetary constraint.
3
Constraint: 2𝑥1 + 3𝑥2 = 18 → 𝑥1 = 9 − 𝑥2
2

3
𝑈 = ln 9 − 𝑥2 + 2ln(𝑥2 )
2
EXERCISES

Exercise 4:
A consumer’s utility function is
𝑈 = ln 𝑥1 + 2ln(𝑥2 )
One good 𝑥1 costs 2 CHF and one good 𝑥2 cost 3 CHF. You can spend a
maximum of 18 CHF on both goods. Find the values of 𝑥1 and 𝑥2 which
maximize U subject to the budgetary constraint.
3
𝑈 = ln 9 − 𝑥2 + 2ln(𝑥2 )
2

3 3 9
𝑑𝑈 − 2 𝑑𝑈 − 𝑥2 + 18 − 3𝑥2 𝑑𝑈 − 𝑥2 + 18
= 2 + =0 → = 2 =0 → = 2 =0
3
𝑑𝑥2 9 − 𝑥 𝑥2 𝑑𝑥2 3 𝑑𝑥2 3
9 − 𝑥2 𝑥2 9 − 𝑥2 𝑥2
2 2 2 2
EXERCISES

Exercise 4:
A consumer’s utility function is
𝑈 = ln 𝑥1 + 2ln(𝑥2 )
One good 𝑥1 costs 2 CHF and one good 𝑥2 cost 3 CHF. You can spend a
maximum of 18 CHF on both goods. Find the values of 𝑥1 and 𝑥2 which
maximize U subject to the budgetary constraint.

9
− 𝑥2 + 18 = 0 𝑥2 = 4
2

3
𝑥1 = 9 − 𝑥2 𝑥1 = 3
2
EXERCISES

Exercise 4:
A consumer’s utility function is
𝑈 = ln 𝑥1 + 2ln(𝑥2 )
One good 𝑥1 costs 2 CHF and one good 𝑥2 cost 3 CHF. You can spend a
maximum of 18 CHF on both goods. Find the values of 𝑥1 and 𝑥2 which
maximize U subject to the budgetary constraint.
9 9 3 9
𝑑𝑈 − 𝑥2 + 18 2
𝑑 𝑈 − 9 − 𝑥2 𝑥2 − − 𝑥2 + 18 (9 − 3𝑥2 )
= 2 =0 2 2 2
3 → 2 = 2
𝑑𝑥2 9 − 𝑥2 𝑥2 𝑑𝑥 2 3
2 9 − 𝑥2 𝑥2
2

𝑑2 𝑈(4)
→ 2 < 0 → 𝑚𝑎𝑥
𝑑𝑥2
EXERCISES

Exercise 5:
You work for a soda company which
produce 330ml cans. Cans are made
of aluminium which is quite expensive.
Your manager asks you to find the
optimal dimensions for a can to use
the least possible aluminium.
(Assume that a can is a perfect cylinder).
EXERCISES

Exercise 5:
You work for a soda company which produce 500ml cans. Cans are made
of aluminium which is quite expensive. Your manager asks you to find the
optimal dimensions for a can to use the least possible aluminium.
(Assume that a can is a perfect cylinder).

𝐴𝑟𝑒𝑎 𝑜𝑓 = 𝜋𝑟 2
𝒓 𝒉
𝑃𝑒𝑟𝑖𝑚𝑒𝑡𝑒𝑟 𝑜𝑓 = 2𝜋𝑟
EXERCISES

𝒓
𝑨𝒓𝒆𝒂 = 𝝅𝒓𝟐
𝒉
𝑽𝒐𝒍𝒖𝒎𝒆 = 𝝅𝒓𝟐 ∗ 𝒉
EXERCISES
EXERCISES

Perimeter of the circle


𝟐𝝅𝒓

𝒉 𝑨𝒓𝒆𝒂 = 𝟐𝝅𝒓 ∗ 𝒉 𝒉

𝑨𝒓𝒆𝒂 = 𝟐𝝅𝒓𝒉 + 𝟐 ∗ 𝝅𝒓𝟐 Area of top and bottom circles


EXERCISES

Exercise 5:
You work for a soda company which produce 330ml cans. Cans are made
of aluminium which is quite expensive. Your manager asks you to find the
optimal dimensions for a can to use the least possible aluminium.
(Assume that a can is a perfect cylinder).

𝑚𝑖𝑛 𝐴𝑟𝑒𝑎 = 2𝜋𝑟ℎ + 2𝜋𝑟 2


𝒓 𝒉
𝑠. 𝑡. 𝑉𝑜𝑙𝑢𝑚𝑒 = 𝜋𝑟 2 ℎ = 0.33𝑙
EXERCISES

Exercise 5:
You work for a soda company which produce 330ml cans. Cans are made
of aluminium which is quite expensive. Your manager asks you to find the
optimal dimensions for a can to use the least possible aluminium.
(Assume that a can is a perfect cylinder).

min 𝐴𝑟𝑒𝑎 = 2𝜋𝑟ℎ + 2𝜋𝑟 2 0.33


𝑠. 𝑡. 𝜋𝑟 2 ℎ = 0.33𝑙 = 0.33𝑑𝑚3 ℎ=
𝜋𝑟 2
0.33 0.66
𝐴𝑟𝑒𝑎 = 2𝜋𝑟 ∗ 2 + 2𝜋𝑟 2 𝐴𝑟𝑒𝑎 = + 2𝜋𝑟 2
𝜋𝑟 𝑟
EXERCISES

Exercise 5:
You work for a soda company which produce 330ml cans. Cans are made
of aluminium which is quite expensive. Your manager asks you to find the
optimal dimensions for a can to use the least possible aluminium.
(Assume that a can is a perfect cylinder).

0.66 𝑑𝐴𝑟𝑒𝑎 −0.66


𝐴𝑟𝑒𝑎 = + 2𝜋𝑟 2 = 2
+ 4𝜋𝑟 = 0
𝑟 𝑑𝑟 𝑟

4𝜋𝑟 3 = 0.66

𝑟 = 0.37 𝑑𝑚 = 37𝑚𝑚
Check that it is a min…
EXERCISES

Exercise 5:
You work for a soda company which produce 330ml cans. Cans are made
of aluminium which is quite expensive. Your manager asks you to find the
optimal dimensions for a can to use the least possible aluminium.
(Assume that a can is a perfect cylinder).

0.33
𝑟 = 0.37 𝑑𝑚 = 37 𝑚𝑚 ℎ= 2
= 0.75 𝑑𝑚 = 75 𝑚𝑚
𝜋𝑟

In reality: 𝑟 = 33 𝑚𝑚 and ℎ = 115 𝑚𝑚


TIMELINE

Last class Today Next class

Chapter 9 Chapter 10 Chapter 11


▪ Partial derivatives ▪ Unconstrained optimization ▪ Integration
▪ Constrained optimization
▪ The substitution method
▪ The Lagrange’s method
QUESTIONS?

You might also like