Professional Documents
Culture Documents
Session 11-12
Session 11-12
Session 11-12
Sessions 11-12
Managerial Implications
11/23/2021 OM 2
What is Forecasting?
• Process of predicting a future event
(The art and science of predicting future events)
• Underlying basis of
all business decisions
• Production
??
• Sourcing
• Inventory
• Personnel
• Facilities
Businesses use forecasts to deal with the uncertainty of the future.
11/23/2021 OM 3
Types of Forecasts
Economic forecasts
◼ Address business cycle – inflation rate,
Demand forecasts
◼ Predict sales of existing or new product/
service
11/23/2021 OM 4
Forecasting Time Horizons
Short-range forecast
◼ Up to 1 year, generally less than 3 months
Long-range forecast
+
◼ 3 years
11/23/2021 OM 7
Influence of Product Life Cycle
Introduction – Growth – Maturity – Decline
Inventory levels
Capacity
11/23/2021 OM 8
Product Life Cycle
Introduction Growth Maturity Decline
11/23/2021 OM 9
Seven Steps in Forecasting
Determine the use of the forecast
Select the items to be forecasted
Determine the time horizon of the forecast
Select the forecasting model(s)
Gather the data
Make the forecast
Validate and implement results
11/23/2021 OM 10
Forecasting Methods
Qualitative: primarily subjective; rely on
judgment and opinion
◼ Used when situation is vague and little data
exist
⚫ New products
⚫ New technology
◼ Involves intuition, experience
⚫ e.g., forecasting sales through m-commerce
11/23/2021 OM 11
Jury of Executive Opinion
• Involves small group of high-level managers
• Relatively quick
• ‘Group-think’
disadvantage
11/23/2021 OM 12
Sales Force Composite
11/23/2021 OM 13
Delphi Method
Iterative group process,
continues until
consensus is reached
3 types of participants
Decision makers
Staff Respondents
(People who can make valuable judgments)
Respondents
11/23/2021 OM 14
Consumer Market Survey
11/23/2021 OM 15
…Forecasting Methods
Quantitative: used when environment is ‘stable’
and historical data exist and is amenable to
mathematical analysis
◼ Existing products
◼ Current technology
1. Naive approach
4. Trend projection
Associative
5. Linear regression Model
11/23/2021 OM 18
Components of Demand
Average demand for a period of time
Trend
Seasonal element
Cyclical elements
Random variation
Autocorrelation
11/23/2021 OM 20
…Components of Demand
Trend
component
Seasonal peaks
Demand for product or service
Actual
demand
Average demand
over four years
Random
variation
| | | |
1 2 3 4
Year
11/23/2021 OM 21
Components of an Observed Demand
Observed demand (O) =
Systematic component (S) + Random component (R)
Level (current deseasonalized demand)
11/23/2021 OM 23
Seasonal Component
Regular pattern of up and down fluctuations
Due to weather, customs, etc.
Occurs within a single year
Number of
Period Length Seasons
Week Day 7
Month Week 4-4.5
Month Day 28-31
Year Quarter 4
Year Month 12
Year Week 52
11/23/2021 OM 24
Cyclical Component
0 5 10 15 20
11/23/2021 OM 25
Random Component
Erratic, unsystematic, ‘residual’
fluctuations
Due to random variation or unforeseen
events
Short duration and
non-repeating
M T W T F
11/23/2021 OM 26
Time Series
A time series is a set of observations
measured at successive points in time or
over successive periods of time. A time
series pattern may have one or more of
the following five characteristics:
✓ Trend
✓ Seasonal
✓ Cyclical
✓ Random Variation
✓ Irregular (one time) Variation
11/23/2021 OM 27
Time Series Forecasting
Quarter Demand Dt
II, 2017 8000
III, 2017 13000
How to
IV, 2017 23000
I, 2018 34000 Forecast
II, 2018 10000 demand
III, 2018 18000 for the
IV, 2018 23000
I, 2019 38000 next four
II, 2019 12000 quarters?
III, 2019 13000
IV, 2019 32000
I, 2020 41000
11/23/2021 OM 28
…Time Series Forecasting
50,000
40,000
30,000
20,000
10,000
0
11/23/2021 OM 29
Time Series
11/23/2021 OM 30
Time Series Forecasting Methods
Goal is to predict systematic component of
demand
◼ Multiplicative: (level)(trend)(seasonal factor)
Static methods
Adaptive methods
11/23/2021 OM 31
…Time Series Forecasting Methods
Static : assume a mixed static model
11/23/2021 OM 33
Moving Average Example
From the example shown earlier: 50,000
40,000
30,000
20,000
10,000
0
Week Demand
1 7 Starting in week 4 and going to
2 9 year 12, forecast demand
3 5 using a 3-week moving
4 9 average; plot the graph.
5 13 Starting in week 4 and going to
6 8 year 12, forecast demand
7 12
8 13
using a 3-week moving
9 9 average; with weights of
10 11 0.1,0.3 and 0.6 plot the graph.
11 7
11/23/2021 OM 35
Moving Average &Weighted Moving Averages
Weighted
30 – moving
average
25 –
Sales demand
20 – Actual
sales
15 –
Moving
10 – average
5 –
| | | | | | | | | | | |
J F M A M J J A S O N D
Figure 4.2
Potential Problems With Moving Average
Increasing n smooths the forecast but
makes it less sensitive to changes
Do not forecast trends well
Require extensive historical data
∑ (weight for period n)
Weighted = x (demand in period n)
moving average ∑ weights
Used when trend is present
Older data usually less important
11/23/2021 OM 37
Weighted Moving Averages: Another Example
Weights Applied Period
3 Last month
2 Two months ago
1 Three months ago
6 Sum of weights
11/23/2021 OM 38
Exponential Smoothing – Problem 4.4
11/23/2021 OM 39
Exponential Smoothing – Problem 4.4
(a) What is the forecast for July?
FJuly= FJune + 0.2 Forecasting Error
FJuly= 42 + 0.2 (40-42) = 41.6
If the centre received 45 million cheques in July,
what would be the forecast for August?
FAugust= FJuly + 0.2 Forecasting Error
FAugust= 41.6 + 0.2 (45-41.6) = 42.3
Why might be this an inappropriate forecasting
method for this situation?
The banking industry may have seasonality in its
processing requirements
11/23/2021 OM 40
Exponential Smoothing– Problem 4.43
Emergency calls for Women helpline for past 24 weeks
are shown in the table below.
Week 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Actual
Value 50 35 25 40 45 35 20 30 35 20 15 40 55 35 25 55 55 40 35 60 75 50 40 65
11/23/2021 OM 42
Least Squares Method
Values of Dependent Variable
Deviation5 Deviation6
^
y = a + bx
Sxy - nxy
b=
Sx2 - nx2
a = y - bx
Least Squares - Example
Time Electrical Power
Year Period (x) Demand x2 xy
2001 1 74 1 74
2002 2 79 4 158
2003 3 80 9 240
2004 4 90 16 360
2005 5 105 25 525
2005 6 142 36 852
2007 7 122 49 854
∑x = 28 ∑y = 692 ∑x2 = 140 ∑xy = 3,063
x=4 y = 98.86
11/23/2021 OM 45
Least Squares - Example
Time Electrical Power
Year Period (x) Demand x2 xy
2009 1 74 1 74
2010 2 79 4 158
2011 3 80 9 240
2012 4 90 16 360
2013 5 105 25 525
2014 6 142 36 852
2015 7 122 49 854
Sx = 28 Sy = 692 Sx2 = 140 Sxy = 3,063
x=4 y = 98.86
Sxy - nxy 3,063 - (7)(4)(98.86)
b= = = 10.54
Sx2 - nx2 140 - (7)(42)
a = y - bx = 98.86 - 10.54(4) = 56.70
Trendline: y^ = 56.70 + 10.54x
11/23/2021 OM 46
Least Squares - Example
130
120 –
110 –
100 –
90 –
80 –
70 –
60 –
50 –
| | | | | | | | |
2001 2002 2003 2004 2005 2006 2007 2008 2009
Year
11/23/2021 OM 47
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 + b x + b x …
1 1 2 2
11/23/2021 OM 48
Seasonality Index– Problem 4.27
Attendance (in thousands) for the newest Disneyland is
given below. Compute seasonal indices.
Quarter Year 1 Year 2 Year 3
Winter 73 65 89
Spring 104 82 146
Summer 168 124 205
Fall 74 52 98
Average
Average Quarterly Seasonal
Quarter 2003 2004 2005 Demand Demand Index Seasonal Indices
0.709
Winter 73 65 89 75.67 106.67 0.709 1.038
Spring 104 82 146 110.67 106.67 1.037
1.553
Summer 168 124 205 165.67 106.67 1.553
Fall 74 52 98 74.67 106.67 0.700
0.700
11/23/2021 OM 49
Measures of Forecast Error
Forecast error = Et = Ft - Dt
s = 1.25MAD
11/23/2021 OM 50
The MAD Statistic to Determine Forecasting Error
n
1 MAD 0.8 standard deviation
A
t=1
t - Ft
1 standard deviation 1.25 MAD
MAD =
n
11/23/2021 OM 51
…Measures of Forecast Error
Mean absolute percentage error (MAPE)
MAPEn = (Sum(t=1 to n)[|Et/ Dt|100])/n
Bias
Shows whether the forecast consistently under- or
overestimates demand; should fluctuate around 0
biasn = Sum(t=1 to n)[Et]
Tracking signal
TSt = bias / MADt
Should be within the range of +6; Otherwise, possibly use a
new forecasting method
11/23/2021 OM 52
MAPE
n
å100 Actual i
- Forecast i / Actuali
MAPE = i=1
n
ACTUAL
TONNAGE FORECAST FOR ABSOLUTE PERCENT ERROR
QUARTER UNLOADED a = .10 100(|ERROR|/ACTUAL)
1 180 175.00 100(5/180) = 2.78%
2 168 175.50 100(7.5/168) = 4.46%
3 159 174.75 100(15.75/159) = 9.90%
4 175 173.18 100(1.82/175) = 1.05%
5 190 173.36 100(16.64/190) = 8.76%
6 205 175.02 100(29.98/205) = 14.62%
7 180 178.02 100(1.98/180) = 1.10%
8 182 178.22 100(3.78/182) = 2.08%
Sum of % errors = 44.75%
MAPE =
å absolute percent error 44.75%
= = 5.59%
n 8
Monitoring and Controlling Forecasts
Tracking Bias
signal =
MAD
∑(actual demand in
period i -
forecast demand
Tracking in period i)
=
signal (∑|actual - forecast|/n)
11/23/2021 OM 54
Tracking Signal
0 MADs Acceptable
range
–
Lower control limit
Time
11/23/2021 OM 55
Tracking Signals
PERCENTAGE OF THE AREA OF THE NORMAL PROBABILITY DISTRIBUTION
WITHIN THE CONTROL LIMITS OF THE TRACKING SIGNAL
11/23/2021 OM 56
Tracking Signal Example
Cumulative
Tracking Absolute Absolute
Signal Forecast
Actual Forecast Forecast
Qtr(Bias/MAD)
Demand Demand Error RSFE Error Error MAD
-10/10 = -1
1
-15/7.5 90
= -2 100 -10 -10 10 10 10.0
2 0/10 =950 100 -5 -15 5 15 7.5
3-10/10115
= -1 100 +15 0 15 30 10.0
4
+5/11 100
= +0.5 110 -10 -10 10 40 10.0
5
+35/14.2125
= +2.5110 +15 +5 15 55 11.0
6 140 110 +30 +35 30 85 14.2
11/23/2021 OM 57
Basic Formula for Adaptive Forecasting
Ft+1 = (Lt + lT)St+1 = forecast for period t+l in period t
Lt = Estimate of level at the end of period t
Tt = Estimate of trend at the end of period t
St = Estimate of seasonal factor for period t
Ft = Forecast of demand for period t (made period t-1 or
earlier)
Dt = Actual demand observed in period t
Et = Forecast error in period t
At = Absolute deviation for period t = |Et|
MAD = Mean Absolute Deviation = average value of At
11/23/2021 OM 58
General Steps in Adaptive Forecasting
1. Initialize: Compute initial estimates of level (L0),
trend (T0), and seasonal factors (S1,…,Sp). This is
done as in static forecasting.
2. Forecast: Forecast demand for period t+1 using the
general equation
3. Estimate error: Compute error Et+1 = Ft+1- Dt+1
4. Modify estimates: Modify the estimates of level
(Lt+1), trend (Tt+1), and seasonal factor (St+p+1), given
the error Et+1 in the forecast
5. Repeat steps 2, 3, and 4 for each subsequent period
11/23/2021 OM 59
Web-Based Forecasting: CPFR
Collaborative Planning, Forecasting, and
Replenishment (CPFR) is a Web-based tool used to
coordinate demand forecasting, production and
purchase planning, and inventory replenishment
among supply chain partners.
Used to integrate the multi-tier or n-Tier supply
chain, including manufacturers, distributors and
retailers.
CPFR’s objective is to exchange selected internal
information to provide for a reliable, longer term
future views of demand in the supply chain.
It uses a cyclic and iterative approach to derive
consensus forecasts.
11/23/2021 OM 60
What is CPFR?
• CPFR is a process used to achieve better
supply chain coordination.
• It is practiced by SC partners for demand
creation and demand fulfillment activities.
• CPFR uses EDI, Internet and other ICT
tools to share information about
o Promotions
o Forecasts
o Item data
o Orders
11/23/2021 OM 61
CPFR as a set of Process and Technology Models
11/23/2021 OM 62
CPFR Business Practices
ANALYSIS PLANNING /
STRATEGY
DEMAND &
EXECUTION SUPPLY
MANAGEMENT
11/23/2021 OM 63
Forecasting in Practice
11/23/2021 OM 64
Popular Methods of Forecasting
Moving
Averages/
Exponential Boosted
Intelligence
Algorithms
Smoothing LSTM Decision
Artificial
Trees
Seq2Seq
Support
ARIMA/ ARFIMA LSTM
Vector
ARCH Regression
Bayesian and Random Lasso
Approach variants Forest Regression GRU Dilated
CNN
1950s early 1970s late 1970s 1980s 1995 1996 1997 1999 2014 2016 2020
11/23/2021 65
Open Source Forecasting Packages
11/23/2021 OM 66
Managerial Implications
Understand the purpose, time horizon, and level of
aggregation in developing a practical forecast .
Be sure to distinguish between demand and sales
Different forecasting methods require different levels of
technical ability and understanding of mathematical
principles and assumptions.
Collaborate in building forecasts
The value of data depends on where you are in the
supply chain
11/23/2021 OM 67
Basic Approach to Demand Forecasting
Understand the objectives of forecasting
Identify major factors that influence the forecast
Integrate demand planning and forecasting
Understand and identify customer segments
Determine the appropriate forecasting technique
Establish performance and error measures for
the forecast
11/23/2021 OM 68
Forecasting in the Service Sector
◼ Unusual events
11/23/2021 OM 69