OM Lec 03

Operations Research & Management
Operations Research
and Management - 4
21MECH42H-O
Dr. Noha Mostafa

Mechanical Engineering department
The British University in Egypt
2
FORECASTING
Dr. Noha Mostafa

Mechanical Engineering department
The British University in Egypt
3
1
Forecasting
What? Why?
Dr. Noha Mostafa

Mechanical Engineering Department
4
“ Forecast:
predict or estimate (a future
event or trend).
”
Dr. Noha Mostafa
5
“He who fails to

plan is planning
to fail”
Sir Winston Churchill Dr. Noha Mostafa
6
Big companies use forecasting
Dr. Noha Mostafa

Operations Research & Management 7
Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

10
2
Forecasting & global
company profile
Dr. Noha Mostafa
Dr. Noha Mostafa

13
 Revenues are derived from people –

how many visitors and how they spend their money.
 Daily management report contains only the forecast and actual attendance at
each park.
 Disney generates daily, weekly, monthly, annual, and 5-year forecasts.
 In Florida park, a staff of 35 analysts and 70 field people survey 1 million

park guests, employees, and travel professionals each year.
Dr. Noha Mostafa

14
 Why?
Forecast used to adjust opening times, rides, shows, staffing levels,
and guests admitted
 Who use it?
Forecast used by labor management, maintenance, operations,
finance, and park scheduling
Dr. Noha Mostafa

15
 Inputs to the forecasting model include airline specials, Federal

Reserve policies, Wall Street trends, vacation/holiday schedules
for 3,000 school districts around the world.
 Average forecast error for the 5-year forecast is 5%.
 Average forecast error for annual forecasts is between 0% and

3%.
Video: Disney data Dr. Noha Mostafa
https://www.youtube.com/watch?v=2iJqSY1ztiQ Mechanical Engineering Department
16
 Revenues in 2018
$20,300,000,000
Dr. Noha Mostafa

Industrial Engineering Department
17
Dr. Noha Mostafa

18
Dr. Noha Mostafa

19
3
Forcasting Time Horizons
Dr. Noha Mostafa
20
From Business Perspective
▰ Forecasting is a planning tool that helps

management in its attempts to cope with the
uncertainty of the future, relying mainly on
data from the past and present and analysis
of trends.
Dr. Noha Mostafa

21
Forecasting time horizons
1. Short-range forecast
▰ Up to 1 year, generally less than 3 months
▰ Purchasing, job scheduling, workforce levels, job assignments, production levels
2. Medium-range forecast
▰ 3 months to 3 years
▰ Sales and production planning, budgeting
3. Long-range forecast
▰ 3+ years
Dr. Noha Mostafa
▰ New product planning, facility location, research and development Mechanical Engineering Department
22
Forecasting time horizons
1. Short-range forecast
▰ Up to 1 year,employs
- usually generallydifferent
less than methodologies
3 months than longer-term forecasting
- tend to job
▰ Purchasing, be more accurate
scheduling, than levels,
workforce longer-term forecastsproduction levels
job assignments,
2. Medium-range forecast
▰ 3 months to 3 years
▰ Sales and production planning, budgeting
Deal with more comprehensive issues and support management
decisionsforecast
3. Long-range regarding planning and products, plants and processes
▰ 3+ years
Dr. Noha Mostafa
▰ New product planning, facility location, research and development Mechanical Engineering Department
23
4
Product Life Cycle
Dr. Noha Mostafa
24
Forecasts helps in:
• Staffing levels
Influence of product life cycle • Inventory levels
• Factory capacity
Introduction Growth Maturity Decline
• Poor time to change

Strategy/Issues
• Best period to increase • Practical to change image, price, or quality

market share price or quality image • Competitive costs become • Cost control critical
• R&D engineering is critical • Strengthen niche critical
• Defend market position
Sales
Video: PLC
25
Influence of product life cycle
Introduction Growth Maturity Decline

26
Dr. Noha Mostafa

27
https://tamersalah.com/coca-cola-business-strategy/
Dr. Noha Mostafa

28
5
Forecasting Approaches
Dr. Noha Mostafa
29
Qualitative methods
▰ Used when situation is vague, and

little data exist
New products, New technology
▰Involves intuition or cultural trend
Dr. Noha Mostafa

30
Quantitative methods
▰Used when situation is ‘stable’ and

historical data exist
Existing products, Current technology
▰Involves mathematical techniques
Forecasting of sales of HD TVs
Dr. Noha Mostafa

31
Possible date sources
▰ Internal
Forecasting data sources based on historical
demand patterns from the company data
▰External
Forecasting data sources based on external International
patterns from information outside the company
Dr. Noha Mostafa
32
6
Overview of Quantitative
Methods
Dr. Noha Mostafa
33
1. Naive approach
Time-Series
2. Moving averages Models
3. Exponential smoothing
4. Trend projection
Associative
5. Linear regression Model
Dr. Noha Mostafa
34
Time-Series Forecasting
▰ A forecasting technique that uses a series of past data

points to make a forecast.
▰Set of evenly spaced numerical data (obtained by
observing response variable at regular time periods).
▰Forecast based only on past values, no other variables
important (assumes that factors influencing past and present
will continue influence in future).
Dr. Noha Mostafa
35
Time-Series Components
Trend Cycles
Seasonality Random
Dr. Noha Mostafa
36
Time-Series Components: Trend
Trend is the gradual upward or downward

Trend movement of the data over time (Typically several
years).
Dr. Noha Mostafa

37
Time-Series Components: Cycles
 Cycles are non-fixed patterns in the data that occur

Cycles every several years (multiple years duration).
 Affected by business cycle, political, and economic
factors.
Dr. Noha Mostafa

0 5 10 15 20 Mechanical Engineering Department
38
Time-Series Components: Seasonality
Seasonality is a data pattern that repeats itself after a

Seasonality fixed period of days, weeks, months, or quarters (within
a single year).
Dr. Noha Mostafa

39
Time-Series Components: Random
 Erratic, unsystematic, ‘residual’ fluctuations

Random  Due to random variation or unforeseen events
 Short duration and non-repeating
Dr. Noha Mostafa

M T W T F
40
Components of demand Trend

component
Demand for product or service
Seasonal peaks
Actual
demand
Average
demand over
Random four years
variation
| | | |
1 2 3 4 Dr. Noha Mostafa
Year Mechanical Engineering Department
41
7
Naive Method
Dr. Noha Mostafa
42
Naïve method
 Assumes demand in next period is the same as

demand in most recent period
e.g., If January sales were 68, then February sales will
be 68
 Sometimes cost effective and efficient
 Can be good starting point

Dr. Noha Mostafa
43
8
Moving Average Method
Dr. Noha Mostafa
44
Moving Average
 MA is a series of arithmetic means
 Used if little or no trend
∑ demand in previous n periods

Moving average = n
Dr. Noha Mostafa
45
Moving Average
Actual 3-Month
Month Shed Sales Moving Average
January 10
February 12
March 13
April 16 (10 + 12 + 13)/3 = 11 2/3
May 19 (12 + 13 + 16)/3 = 13 2/3
June 23 (13 + 16 + 19)/3 = 16
July 26 (16 + 19 + 23)/3 = 19 1/3
Dr. Noha Mostafa

46
Moving Average
Moving Average
Forecast
30 –
28 –
Actual Sales
26 –
24 –
Shed Sales
22 –
20 –
18 –
16 –
14 –
12 –
10 –
| | | | | | | | | | | |
J F M A M J J A S O N D
Dr. Noha Mostafa
47
Weighted Moving Average
 Used when trend is present (Older data usually less important)
 Weights based on experience and intuition
∑ (weight for period n)

Weighted x (demand in period n)
moving average = ∑ weights
Dr. Noha Mostafa
48
Actual 3-Month Weighted

January 10
February 12
March 13
April 16 [(3 x 13) + (2 x 12) + (10)]/6 = 121/6
May
Weights Applied 19 x 16) + (2 x 13) + (12)]/6 = 141/3
[(3Period
June 3 23 [(3 x 19) + (2 x 16) + (13)]/6 = 17
Last month
July 2 26 [(3 xago
Two months 23) + (2 x 19) + (16)]/6 = 201/2
1 Three months ago
6 Sum of weights Dr. Noha Mostafa
49
Actual 3-Month Weighted

January 10
February 12
March 13
April 16 [(3 x 13) + (2 x 12) + (10)]/6 = 121/6
May 19 [(3 x 16) + (2 x 13) + (12)]/6 = 141/3
June 23 [(3 x 19) + (2 x 16) + (13)]/6 = 17
July 26 [(3 x 23) + (2 x 19) + (16)]/6 = 201/2
Dr. Noha Mostafa

50
Problems with MAs
 Increasing the number of periods (n)

smooths the forecast but makes it less
sensitive to changes
 Do not forecast trends well
 Require extensive historical data
Dr. Noha Mostafa

51
MA and WMA
Weighted
30 – moving
average
25 –
Sales demand
20 – Actual
sales
15 –
Moving
10 – average
5 – | | | | | | | | | | | | Dr. Noha Mostafa

J F M A M J J A S O N D Mechanical Engineering Department
52
9
Exponential Smoothing
Dr. Noha Mostafa

53
Exponential smoothing
 Form of weighted moving average

 Weights decline exponentially
 Most recent data weighted most
 Requires smoothing constant (a)
 Ranges from 0 to 1
 Subjectively chosen
 Involves little record keeping of past data Dr. Noha Mostafa
54
New forecast = Last period’s forecast + α (Last period’s actual

demand – Last period’s forecast)
Ft = Ft – 1 + α(At – 1 - Ft – 1)
Where
Ft = new forecast
Ft – 1 = previous forecast
α = smoothing (or weighting) constant (0 ≤ α ≤ 1)

Dr. Noha Mostafa
55
▰Predicted demand for February = 142 Ford Mustangs

▰Actual demand in February = 153
▰Smoothing constant a = 0.20
▰What is the forecast for March demand?
Ft = Ft – 1 + α(At – 1 - Ft – 1)
March forecast = 142 + 0.2(153 – 142)

Dr. Noha Mostafa
≈ 144 cars Mechanical Engineering Department
56
10
Forecasting Error
Dr. Noha Mostafa
57
Forecast error
 The objective is to obtain the most accurate forecast no matter the

technique
 We generally do this by selecting the model that gives us the lowest
forecast error
▰ Forecast error = Actual demand - Forecast value = At - Ft
Dr. Noha Mostafa

58
Common Measures of Error
Mean Absolute Deviation (MAD): A measure of the overall forecast error for a model.
∑ |Actual - Forecast|
MAD = n
Mean Squared Error (MSE): The average of the squared differences between the forecasted and
observed values.
∑ (Forecast Errors)2
MSE = n
Dr. Noha Mostafa
59
11
Associaive Forecasting
Dr. Noha Mostafa
60
When to use associative forecasting?
 Unlike time-series forecasting, associative forecasting models usually

consider several variables that are related to the quantity being
predicted.
 Once these related variables have been found, a statistical model is
built and used to forecast the item of interest.
 This approach is more powerful than the time-series methods that use
only the historic values for the forecasted variable.
Dr. Noha Mostafa
61
iPhone sales Advertising

budget
Competitors’
prices
Company’s prices
Promotional
strategies
Dr. Noha Mostafa

62
Linear regression analysis
 Most common technique is linear regression analysis.

 Forecasting an outcome based on predictor variables using the least squares
technique.
^
y = a + bx
where
^y = computed value of the variable to be predicted (dependent variable)
a = y-axis intercept
b = slope of the regression line
x = the independent variable though to predict the value of the
Dr. Noha Mostafa
dependent variable Mechanical Engineering Department
63
Example
Nodel construction company
Sales Local Payroll

($ millions), y ($ billions), x
2.0 1
3.0 3
2.5 4
2.0 2
2.0 1
3.5 7
Dr. Noha Mostafa

64
Example
 The company wants to determine whether there is a linear relationship between

sales and area payroll.
4.0 –
3.0 –
Sales
2.0 –
1.0 –
| | | | | | |
0 1 2 3 4 5 6 7
Area payroll Dr. Noha Mostafa
65
Example
Sales, y Payroll, x x2 xy
2.0 1 1 2.0
3.0 3 9 9.0
2.5 4 16 10.0
2.0 2 4 4.0
2.0 1 1 2.0
3.5 7 49 24.5
∑y = 15.0 ∑x = 18 ∑x2 = 80 ∑xy = 51.5
∑xy - nxy 51.5 - (6)(3)(2.5)

x = ∑x/6 = 18/6 = 3 b= = = 0.25
∑x2 - nx2 80 - (6)(32)
y = ∑y/6 = 15/6 = 2.5 a = y - bx = 2.5 - (0.25)(3) = 1.75 Dr. Noha Mostafa
66
Example
^
y = 1.75 + 0.25x Sales = 1.75 + 0.25(payroll)
4.0 – If payroll next year is estimated to be $6

billion, then:
3.25
3.0 – Sales = 1.75 + .25(6)
Sales
Sales = $3,250,000
2.0 –
1.0 –
| | | | | | |
0 1 2 3 4 5 6 7
Area payroll Dr. Noha Mostafa
67
Correlation
 How strong is the linear relationship between the two variables?
 Correlation does not necessarily imply causality!
 Coefficient of correlation, r, measures degree of association
 Values of r range from -1 to +1
Dr. Noha Mostafa

68
Correlation coefficient
Dr. Noha Mostafa

69
Coef. of correlation
y y
(a) Perfect positive x (b) Positive x

correlation: correlation:
r = +1 0<r<1
y y
(c) Perfect negative x (d) No correlation: x Dr. Noha Mostafa

correlation: r=0
r = -1 Mechanical Engineering Department
70
Coef. of determination
 Coefficient of Determination, r2, measures the percent of variation in
y that is explained by the regression equation
 Values range from 0 to 1
 Easy to interpret
Dr. Noha Mostafa

71
Example
Nodel construction company
Sales Local Payroll

($ millions), y ($ billions), x
2.0 1
3.0 3
2.5 4
2.0 2
2.0 1
3.5 7
Dr. Noha Mostafa

72
Example
2.0 1 1 2.0
3.0 3 9 9.0
2.5 4 16 10.0
2.0 2 4 4.0
2.0 1 1 2.0
3.5 7 49 24.5
∑y = 15.0 ∑x = 18 ∑x2 = 80 ∑xy = 51.5
∑y2 = 39.5 r=? r2 = ?

Dr. Noha Mostafa
73
Example
2.0 1 1 2.0
3.0 3 9 9.0
2.5 4 16 10.0
2.0 2 4 4.0
2.0 1 1 2.0
3.5 7 49 24.5
∑y = 15.0 ∑x = 18 ∑x2 = 80 ∑xy = 51.5
∑y2 = 39.5 r = 0.901 r2 = 0.81

Dr. Noha Mostafa
74
Multiple Regression Analysis
 If more than one independent variable is to be used in the model, linear

regression can be extended to multiple regression to accommodate
several independent variables.
y^ = a + b1x1 + b2x2 …
 Computationally, this is quite complex and generally done on the

computer.
Dr. Noha Mostafa
75
Multiple regression
 In the Nodel example, including interest rates in the model gives the
new equation:
^
y = 1.80 + .30x1 - 5.0x2
 An improved correlation coefficient of r = 0.96 means this model

does a better job of predicting the change in construction sales.
Dr. Noha Mostafa

76
10
Qualitative forecasting
Dr. Noha Mostafa
77
 Using expert opinion and collective experience to unlock the secrets
of the future.
 The company prepares its sales forecasts by asking knowledgeable
people such as experts in the field, sales personnel, customers, and
company executives.
Dr. Noha Mostafa
78
When to use qualitative forecasting:
 Data as an historical series is not available, or when the planning horizon
is very long.
 An unusual product or a unique project is being contemplated.
 It can be also used to adjust the quantitatively generated forecast.
 When markets have been disrupted by strikes, wars, natural disasters,

Dr. Noha Mostafa
recessions, or inflation. Mechanical Engineering Department
79
Quan. Vs Qual.
Dr. Noha Mostafa

80
Techniques and tools
 Survey
 Groups
 Market research
Dr. Noha Mostafa

81
Survey
 Data can be gathered by phone or in writing.

 Data comes in three categories:
1. Highly valuable
2. Absolutely essential
3. Supporting material.
 The survey group is known as the ‘reference population’.
Dr. Noha Mostafa
82
Group techniques
 Jury of Executive Opinion,

 Delphi Method,
 Sales Force Composite,
 Supply Chain Partner Forecasting,
 Grassroots forecasting,
 Nominal group,
 the dialectical inquiry,
 Leading indicators Dr. Noha Mostafa
83
Jury of Executive Opinion
 Opinions from a small group of company

executives (senior managers) that meets
together to predict sales.
 The executives’ predictions are averaged so
that the forecast is a composite of their points
of view.
Dr. Noha Mostafa
84
Jury of Executive Opinion
 Advantages:
 Based on reliable inside opinion
 Quick and easy to use
 Disadvantages:
 Results depend on executives’ skills
 All predictions carry equal weight.
 Group thinking, personality dominance. Dr. Noha Mostafa
85
Delphi method (The expert survey)
The Oracle of Delphi
Dr. Noha Mostafa

86
 It involves company executives and outside experts such as university

professors, consultants, or industry analysts (5-10 experts).
 Predictions are made secretly and then averaged together. The results of
the first poll are sent to the experts, who are asked to respond with a
second opinion. The process is repeated until a very narrow, firm consensus
is agreed upon.
Dr. Noha Mostafa
87
Dr. Noha Mostafa

88
 Advantages:
 Can prevent social pressure and group thinking
 Can prevent forceful individuals from dominating others
 Can prevent time-consuming discussions or arguments
 Can gather opinions from those who won’t speak out in groups
 Disadvantages:
 Takes a lot of time to complete multiple rounds of the process
 Can be expensive Dr. Noha Mostafa
89
Sales Force Composite
 This method gathers opinions from

the sales force.
 Each salesperson forecasts his/her
sales for a future period in his
territory.
 The sales analyst then adds those
forecasts together to get the
sales force composite forecast for
the period.
Dr. Noha Mostafa
90
 Advantages:
 Accurate forecasts for individual products (The sales force works directly with
customers and understands the demand for certain products.)
 Higher sales totals (When the sales force predicts its own sales, sales personnel
are more motivated to achieve those numbers.)
 Inexpensive to use
 Provides detailed information
Dr. Noha Mostafa
91
 Disadvantages:
 Lacks a long-range view (The sales force may not have enough information
about the company’s future plans to accurately predict long- term sales.)
 Sales force resentment due to having to take time away from selling to
prepare sales forecasts.
 Forecasts that benefit sales force (may forecast sales lower than s/he thinks
can be achieved to be sure the forecast is met.)
Dr. Noha Mostafa
92
Market research
 Marketing research is a qualitative method based on the consumers’

buying intentions.
 It is a systematic gathering, recording and analyzing of data about
customers and competitors relating to the marketing of goods and
services
https://www.zdnet.com/article/coronavirus-to-test-supply-chain-analytics-business-demand-visibility/
https://www.idc.com/getdoc.jsp?containerId=US46208420 Dr. Noha Mostafa
https://www.altexsoft.com/blog/demand-forecasting-methods-using-machine-learning/
https://www.oliverwyman.com/our-expertise/insights/2020/nov/machine-learning-will-revolutionize-forecasting.html Mechanical Engineering Department
THANKS!
Any questions?
You can find me at
Noha.Mostafa@bue.edu.eg

OM Lec 03

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

OM Lec 03

Uploaded by

Copyright:

Available Formats

Operations Research & Management

Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

“He who fails to

Big companies use forecasting

Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

 Revenues are derived from people –

 Disney generates daily, weekly, monthly, annual, and 5-year forecasts.

 In Florida park, a staff of 35 analysts and 70 field people survey 1 million

Dr. Noha Mostafa

Dr. Noha Mostafa

 Inputs to the forecasting model include airline specials, Federal

 Average forecast error for the 5-year forecast is 5%.

 Average forecast error for annual forecasts is between 0% and

Dr. Noha Mostafa

Dr. Noha Mostafa

Dr. Noha Mostafa

From Business Perspective

▰ Forecasting is a planning tool that helps

Dr. Noha Mostafa

Forecasting time horizons

Forecasting time horizons

Introduction Growth Maturity Decline

• Poor time to change

• Best period to increase • Practical to change image, price, or quality

Influence of product life cycle

Introduction Growth Maturity Decline

Influence of product life cycle

Dr. Noha Mostafa

Influence of product life cycle

Dr. Noha Mostafa

▰ Used when situation is vague, and

Dr. Noha Mostafa

▰Used when situation is ‘stable’ and

Dr. Noha Mostafa

Possible date sources

▰ A forecasting technique that uses a series of past data

Time-Series Components: Trend

Trend is the gradual upward or downward

Dr. Noha Mostafa

Time-Series Components: Cycles

 Cycles are non-fixed patterns in the data that occur

Dr. Noha Mostafa

Time-Series Components: Seasonality

Seasonality is a data pattern that repeats itself after a

Dr. Noha Mostafa

Time-Series Components: Random

 Erratic, unsystematic, ‘residual’ fluctuations

Dr. Noha Mostafa

Components of demand Trend

 Assumes demand in next period is the same as

 Sometimes cost effective and efficient

 Can be good starting point

 MA is a series of arithmetic means

 Used if little or no trend

∑ demand in previous n periods

Dr. Noha Mostafa

 Used when trend is present (Older data usually less important)