Chapters 4.6
Optimization
Optimization
is using calculus to either
•
minimize
an undesirable quantity, such as
cost, or
•
maximize
a desirable quantity, such as
profit, revenue or demand.
In this section we pursue optimization of
narrative problems.
First, however, we would like to review some
derivative rules and examples.
Derivatives of functions requiring arguments:
Example:
Example:
Generalized power rule:
If
then
Product Rule:
Quotient Rule:
In Chapter 4.4 we indicated that a
point of
diminishing returns
occurs at the
inflection
point
of an increasing function which changes
concavity from upwards to downwards, as shown
in the graph below.
At the inflection
is
increasing most
rapidly
and the rate of
return for the investment
is
optimized
.
S(a)
Problem 1:
Telemarketing is the major means of advertising
for a small business. The following table provides
the monthly profit of the business, in thousands of
dollars, as a function of monthly labor hours
devoted to telemarketing.
telemarketing labor hours
profit ($Th) telemarketing
labor hours
profit ($Th)
0 2 50 43.0
10 3.5 60 48.5
20 8.5 70 55.5
A. Scatterplot the telemarketing data. You should
find that a cubic model would be appropriate.
Determine the cubic model, reporting the model
parameters and R
2to 5 decimal digits.
B. Find the 1
stand 2
ndderivative of the cubic
model.
C. For what number of telemarketing labor hours is
profit increasing most rapidly?
D. What is the profit associated with your result
from Part C?
Problem 1 Solved:
A.
Problem 1 Solved (cont):
B.Problem 1 Solved (cont):
C.an inflection point occurs at x = 39.3
Since the model is an increasing function, the inflection point indicates the input value at which the output is
increasing most rapidly. That is, when 39.3 labor hours are invested in telemarketing, profit is increasing most rapidly.
This is the point of diminishing returns. Investing more hours into telemarketing will generate more profit, but at a diminishing rate.
Problem 1 Solved (cont):
D. an inflection point occurs at x = 39.3
Monthly profit is approximately $29,290 when 39.3 hours are invested in telemarketing monthly.
Problem 1 Solved (cont):
E. an inflection point occurs at x = 39.3
Profit is increasing at approximately $1,060 per hour when 39.3 hours are invested in telemarketing monthly.
Consider:
Market share
is the percentage of a market (defined in
terms of either units or revenue) accounted for by a
specific company or organization. Market share is a
key indicator of market competitiveness—that is, how
well a firm is doing against its competitors.
The market share for a company’s product, , months
since its introduction into the marketplace, can be
modeled with the following function during the first year
after its introduction:
Problem 2:
A.
Find the first derivative of the market share
model.
B.
Find the second derivative of the market
share model.
A.
Recall the quotient rule:
B. Starting with the first derivative and using the quotient rule:
Problem 2 Solved:
(
� 1� 2
)
′
(�)= �1
′
(�) � 2(�)− �2′(�) �1(�)
Problem 2:
C.
Does the model reach a relative extreme
over the interval 0
t
12, that is, during
the first year.
D.
When did the product reach its maximum
market share during the first year and
C.
relative extreme occurs at months
D. candidate input values at which a maximum market share occurs on the interval :
t = 3, 0, 12
maximum market share is 22%, and occurs 3 months after product introduced
Problem 2 Solved (cont):
Problem 2:
E.
Graph the market share function over the
first year.
F.
Your graph should indicated the
rate of
change
of the market share reaches a
maximum negative value around 4 to 6
months after introduction. Thereafter, the
rate of change begins to slow. Determine
the time when the rate of change of
E. See graph below F.
using the solver, , 5.2, -5.2
Decline in market share begins to slow 5.2 months after introduction of the product.
Problem 2 Solved (cont):
Consider:
A Greyhound bus route currently transports, on
average, 8000 passengers per month at a fare of $50.
Market research indicates the company will lose 100
monthly passengers for each $1 increase in fare.
We wish to determine the fare which maximizes
revenue.
The number of passengers per month, N, and the
revenue, R, will be dependent upon the fare. Therefore
identify:
The statement “the company will lose 100 monthly
passengers for each $1 increase in fare” indicates the
rate of change
in number of passengers, with respect
to the fare, will be .
This is a
constant rate of change
and indicates the
relationship between number of passengers and the
fare is
linear
. That is, a straight line can express the
relationship between fare and number of passengers:
Recall:
x = fare per passenger ($/passenger)
N(x) = number of passengers per month
So far we know
The statement “8000 passengers per month at a fare of
$50” gives an ordered pair:
Solving for b:
The revenue, , is the fare/passenger, , times the
number of passengers, .
Revenue may be maximized for the fare,
, which yields
a zero derivative and negative 2
ndderivative:
Problem 3:
A small ski resort offers an end-of-season $200 weekend special providing it can sell at least 100 reservations. The maximum
number of reservations is the facility capacity of 160. For each reservation in excess of 100, the price will be decreased by $1 per reservation.
We wish to determine:
• the number of reservations which maximizes revenue, and • the maximum revenue.
The price per reservation, P, and the revenue, R, are dependent upon the number of reservations. Identify:
x = number of reservations
P(x) = price per reservation ($/reservation) R(x) = revenue ($)
Problem 3 (cont):
The price per reservation, P, and the revenue, R, are dependent upon the number of reservations. Identify:
x = number of reservations
P(x) = price per reservation ($/reservation) R(x) = revenue ($)
Note:
1. We are considering price, , as a function of the number of reservations, .
2. Revenue is the product of the number of reservations and the price per reservation, that is,
3. Note that in this problem we are constrained by 100 ≤ x ≤ 160.
Problem 3 (cont):
A. The statement “For each reservation in excess of 100, the price will be decreased by $1 per reservation” indicates the
rate of change in price per reservation, with respect to the
number of reservations, will be . This is a constant rate of
change and indicates the relationship between number of
reservations and the price is linear. That is, a straight line can express this relationship between the input variable, x, and the output variable, P(x). Write a generic linear model relating these two variables.
B. What is the value of the slope of this linear relationship?
C. The statement “$200 weekend special providing it can sell at least 100 reservations” gives an ordered pair: . Use this
statement to find the vertical axis intercept, b, of the linear model.
A.
B.
C. one ordered pair: Solving for b:
D.
Problem 3 (cont):
E. Write the equation for R(x).
F. Does R(x) reach a relative extreme on the interval 100 ≤ x ≤ 160?
G. Determine the 2nd derivative of the revenue function and indicate what it tells us about any relative extreme values.
H. Evaluate all candidate input values to determine the maximum revenue.
E.
F.
G. any relative extreme is a max
produces a relative maximum revenue
Problem 3 Solved (cont):
H. Candidate number of reservations, x, which may maximize revenue on the interval 100 ≤ x ≤ 160:
produces a maximum revenue of $22,500
a price of $150 maximizes revenue.
Problem 3 Solved (cont):
Consider:
Frozen orange juice sells consistently thru the year. A
supermarket wishes to optimize the ordering of OJ by
submitting orders of the same size thru the year. It is
estimated annual sales will be 1200 cases.
The manager is concerned with two competing costs:
First, each delivery will cost $75, so less deliveries per
year reduces the delivery costs.
However, the storage cost per case is $8 per year
where the average number of cases stored per year is
50% of the average number of cases per order. Less
deliveries means more cases per order, increasing
Set
= # of annual orders
Since we require 1200 cases annually, the number of
cases per order is:
= # cases/order
Delivery costs, , are a function of the # of orders,
specifically, $75 per order:
Storage costs, , are a function of the # cases per
order, . Specifically, 50% of the cases per order, are
expected to be stored per year at an annual cost of $8
per case:
Problem 4:
A.Write the total cost function, .
B.Determine the number of orders, , which minimizes
the total cost function.
A. , x = # of annual orders
B.
which is always greater than zero
Therefore produces a relative minimum cost C.
Problem 5:
A tire manufacturer has annual orders for 600,000 for a specific tire model. The production costs, Cp, to set up the plant for each model are $15,000 per production run.
Storage cost, Cs, per tire per year are $5. Assume the average number of tires stored per year is 50% of the average number of tires per run.
Again, less production runs reduce production costs at the expense of storage costs.
Let = # of tires produced per run
= # of tires produced per run
Since we require 600,000 tires annually: = # runs
Production costs, , are a function of the # of runs, specifically, $15,000 per run:
Storage costs, , are a function of the # tires per run, .
Specifically, 50% of the tires per run, are expected to be stored per year at an annual cost of $5 per tire:
A.
> 0 therefore the x value generates a relative minimum cost B.
C.
Problem 6:
A museum experiments with the Sunday ticket price,
generating the following data.
ticket price ($)
average # Sunday
customers
20
1600
25
1500
30
1250
32
850
A.Scatterplot the data and find the best model
which
expresses S(x), that is, # of customers as a function
of ticket price, x.
B.Write an expression for revenue, R(x), as a
function of ticket price, x. Note, R(x) = x S(x).
C.Determine the ticket price which maximizes
revenue over the interval $20 ≤ x ≤ $32.
D.The museum estimates that the average total cost
per customer to operate the facility is $10. Write an
expression for cost, C(x), as a
function of ticket
price, x. Note, C(x) = $10 S(x).
A. B.
C.
relative minimum revenue
relative maximum revenue
Max revenue on interval $20 ≤ x ≤ $32:
D. E.
relative min
relative max
E. Max profit on interval $20 ≤ x ≤ $32:
The ticket price which maximizes profit, $27.58, is a bit different than that which maximizes revenue, $26.05.