Mba om 05_forecasting_foroperations

Operations Management
Author: Prof. Niranjana K.R.
B.E. (Mech), PGDM, SSBB, LA ISO9001 & AS9100, Member – PMI & QCFI
Email: niranjanakoodavalli@gmail.com

Agreements
24-08-2016 Author: Niranjana K R 2

References
1. Modern Production/Operations Management, - Elwood S.
Buffa, Rakesh K.Sarin, 8th Edition, WILLY
2. Operations Management Theory and Practice, 2e - B.
Mahadevan
3. Operations Management For Competitive Advantage – 11th
edition – Richard B. Chase, F. Robert Jacobs, Nicholas J.
Aquilano
4. Operations Management – An Integrated Approach, by R. Dan
Reid, Nada R. Sanders, 5th Edition, Wiley ,2012.
5. Operations Management, 10th edition Author(s): Jay Heizer,
Barry Render

Operations Management
Forecasting for Operations
Author: Prof. Niranjana K.R.
B.E. (Mech), PGDM, SSBB, LA ISO9001 & AS9100, Member – PMI & QCFI
Email: niranjanakoodavalli@gmail.com

Learning Objectives
• Identify Principles of Forecasting
• Explain the steps in the forecasting process
• Identify types of forecasting methods and their
characteristics
• Describe time series and causal models
• Generate forecasts for data with different patterns:
level, trend, seasonality, and cyclical
• Describe causal modeling using linear regression
• Compute forecast accuracy
• Explain how forecasting models should be selected
Author: Niranjana K R24-08-2016 5

Previous Exam Questions
1. Explain the different types of forecasting and give their merits and
demerits? – MU Essay Question, June/July 2012 (2011 scheme)
2. Discuss the process of facilities management. (a) Demand forecast. –
Mar/Apr 2012 (2008 scheme)
3. Enumerate the importance of accuracy in forecasting? Briefly explain the
criteria used in selecting the best forecasting methods.- Aug 2014
(2007/2009)

Predict the next number in the pattern
a) 3.7, 3.7, 3.7, 3.7, 3.7, ?
b) 2.5, 4.5, 6.5, 8.5, 10.5, ?
c) 5.0, 7.5, 6.0, 4.5, 7.0, 9.5, 8.0, 6.5, ?
24-08-2016 Author: Niranjana K R
3.7
12.5
9.0
Forecasting means Predicting future events.
7

Principles of Forecasting
There are many types of forecasting models that differ in
complexity and amount of data & way they generate forecasts:
1. Forecasts are rarely perfect
2. Forecasts are more accurate for groups or families of items
rather than for individual items.
3. Forecasts are more accurate for shorter than longer time
horizons.

Why do we Forecast?
• Forecasting activity typically precedes a planning process.
• Forecasting plays an important role as a tool for the planning
process.
• Applications of Forecasting – Next slide

Applications of Forecasting (1 of 2)
• Dynamic and complex environments
– Sales forecast
• Short-term fluctuations in production
– Handle short-term demand fluctuations
– Avoid knee-jerk reactions
• Better materials management
– Ensure better material and greater availability of resources
Continued…

Applications of Forecasting (2 of 2)
• Rationalized manpower decisions
– Nature of resources, their timing and magnitude
• Basis for Planning and Scheduling
– Better Planning and Scheduling
• Strategic decisions
– Planning for product line,
– New products,
– Augmenting capacity,
– Building new factories,
– Expansion of business etc.

Forecasting Time Horizon - Implications
Criterion Short-term Medium-term Long-term
Typical Duration 1-3 Months 12-18 Months 5-10 Years
Nature of Decisions Purely tactical Tactical as well as
strategic
Purely strategic
Key considerations Random (short-term)
effects
Seasonal and cyclical
effects
Long-term trends and
business cycles
Nature of data Mostly quantitative Subjective and
quantitative
Largely subjective
Degree of Uncertainty Low Significant High
Examples • Revising quarterly
production plans
• Rescheduling supply of raw
materials
• Annual production
planning
• Capacity augmentation
• New Product Introduction
• Facilities location decisions
• New business development

Forecasting During the Life Cycle
Introduction Growth Maturity Decline
Sales
Time
Quantitative models
- Time series analysis
- Regression analysis
Qualitative models
- Executive judgment
- Market research
-Survey of sales force
-Delphi method
24-08-2016 13Author: Niranjana K R

Steps in the Forecasting Process
1. Decide what needs to be forecast
– Level of detail, units of analysis & time horizon required
2. Evaluate and analyze appropriate data
– Identify needed data & whether it’s available
3. Select and test the forecasting model
– Cost, ease of use & accuracy
4. Generate the forecast
5. Monitor forecast accuracy

Steps in Developing the forecasting logic - Flow chart
24-08-2016 Author: Niranjana K R
Start
Stop
Identify:
1. The purpose of the forecast
2. The time horizon
3. The type of data needed
Identify a Technique:
1. Collect/Analyze past data
2. Select an appropriate model
Develop the Forecasting Logic:
1. Establish model parameters
2. Build the model
Test the model adequacy using
historical data
Satisfactory
15

Sources of data
• Sales-force Estimates
• Point of Sales (POS) Data Systems
• Forecasts from Supply Chain Partners
• Trade/Industry Association Journals
• B2B Portals/Marketplaces
• Economic Surveys and Indicators
• Subjective knowledge

Basic Categories of Forecasting Methods (1 of 4)
• Forecasting methods can be divided into three main
categories:
– Extrapolative or time series methods – Make use of past
data to prepare future estimates.
– Causal or explanatory methods – Analyze the data from
the viewpoint of a cause-effect relationship.
– Qualitative or judgmental methods – Judgmental methods
rely on experts’ opinion.
• In some situations, a combination of methods may
be more appropriate than a single method.

• Extrapolative or time series methods:
– Make use of past history of demand in making a
forecast for the future.
– The objective is to identify the pattern in historic
data and extrapolate this pattern for the future.
– Very similar to driving while looking only through
a rear view mirror
– This method works well when the time horizon for
which the forecast is made is short

• Causal or explanatory methods
– Analyze the data from the viewpoint of a cause-effect
relationship.
– Causal methods of forecasting assume that the demand for
an item depends on one or more independent factors
(e.g., price, advertising, competitor’s price, etc.)
– These methods seek to establish a relationship between
the variable to be forecasted and independent variables.
– Once this relationship is established, future values can be
forecasted by simply plugging in the appropriate values for
the independent variable.

• Qualitative or judgmental methods
– Judgmental methods rely on experts’ opinion in
making a prediction for the future.
– Useful for medium to long-range forecasting tasks.
– Sounds unscientific and ad hoc.
– Very useful method when,
• Past data are unavailable or not representative of the
future,
• There are few alternatives other than using the
informed opinion of knowledgeable people.

Summary of Forecasting Methods

Forecasting Approaches
Qualitative Methods – Judgmental methods
• Forecast is made subjectively by the forecaster
• They are educated guesses based on intuition, knowledge, and
experience.
• Used when situation is vague and little data exist
– New products
– New technology
• Involves intuition, experience
– e.g., forecasting sales on Internet
• Forecasts are biased as made by people.

Qualitative Methods - Summary
Author: Niranjana K R 2324-08-2016

Quantitative Methods
• Quantitative methods are Based on Mathematics
• Divided into two categories - Time Series models and Causal models
• Time Series Models:
– A time series is a series of observations taken at regular intervals over a specified
period of time.
• Ex: If you were forecasting quarterly corporate sales and had collected five years of
quarterly sales data, you would have time series.
– Assumes information needed to generate a forecast is contained in a time series
of data
– Assumes the future will follow same patterns as the past
• Causal Models or Associative Models
– Explores cause-and-effect relationships
– Uses leading indicators to predict the future
– Housing starts and appliance sales

Quantitative Forecasting Models (1 of 2)

Overview of Selected Quantitative
Approaches
1. Naive approach
2. Moving averages
3. Exponential smoothing
4. Trend projection
5. Linear regression
time-series
models
associative
model

Extrapolative Methods
• Components of Demand
– The horizontal component
• This type of demand exists when the demand fluctuates about an average
demand.
• The average demand remains constant and does not consistently increase
or decrease
• The sales of a product in the mature stage of the product lifecycle may
show a horizontal demand pattern.
– The trend component
– The seasonal component
– The cyclical component
– The Random component
Trend
Seasonal
Cyclical
Random

Components of Demand
Author: Niranjana K R
Demandforproductorservice
| | | |
1 2 3 4
Time (years)
Average demand
over 4 years
Trend
component
Actual demand
line
Random variation
Figure 4.1
Seasonal peaks
24-08-2016 29

Trend Component
• Refers to a sustained increase or decrease in demand from one period to
the next.
– Example: If the avg. monthly demand for a product has increased 10 to 15 % in each
of the past few years, then an upward trend of demand exists.
• Persistent, overall upward or downward pattern
• Changes due to population, technology, age, culture, etc.
• Typically several years duration

Seasonal Component
• Influenced by the seasonal factors that impact demand positively or
negatively
 Example: The sales of snow blowers will be higher in winter months and lower in
summer months.
• Regular pattern of up and down fluctuations
• Due to weather, customs, etc.
• Occurs within a single year
Period Length Number of Seasons
Week Day 7
Month Week 4-4.5
Month Day 28-31
Year Quarter 4
Year Month 12
Year Week 52

Cyclical Component
• Repeating up and down movements
• Affected by business cycle, political, and economic factors
• Multiple years duration
• Often causal or associative relationships
0 5 10 15 20

Random Component
 Random variations are unexplained variations that cannot be
predicted
 Erratic, unsystematic, ‘residual’ fluctuations
 Due to random variation or unforeseen events
 Short duration and non-repeating
Data= level + trend + seasonality + cycles + random variation
Author: Niranjana K R M T W T F24-08-2016 33
Pattern

Time Series Models
• Forecaster looks for data patterns as
– Data = historic pattern + random variation
• Historic pattern to be forecasted:
– Level (long-term average) – data fluctuates around a constant mean
– Trend – data exhibits an increasing or decreasing pattern
– Seasonality – any pattern that regularly repeats itself and is of a
constant length
– Cycle – patterns created by economic fluctuations
• Random Variation cannot be predicted

• Naive: The forecast is equal to the actual value observed
during the last period – good for level patterns
Ft+1 = At
Where, Ft+1 = forecast for next period, t+1, At = actual value for current period, t
Naive
A restaurant is forecasting sales of chicken dinners for the month of April. Total sales
of chicken dinners for March were 320. If management uses the naïve method to
forecast, what is their forecast of chicken dinners for the month of April?
Solution: Our equation is Ft+1 = At
Adding the appropriate time period:
F April = A March
F April = 320 dinners

Forecasting: Time Series Models

Simple Mean or Average
• One of the simplest averaging models.
• The forecast is made by simply taking an average of all data:
Ft+1 = Σ At/n = At +At-1 + ……. + At-n /n
Where, Ft+1 = forecast of demand for next period, t+1, At = actual value
for current period, t and n= number of periods or data points to be
averaged

Problem: Simple Mean or Average:
New Tools Corporation is forecasting sales for its classic product, Handy-
Wrench. Handy-Wrench sales have been steady, and the company uses a
simple mean to forecast. Weekly sales over the past five weeks are available.
Use the mean to make a forecast for week 6.
Time Period
(in weeks)
Actual Sales Forecast
1 51
2 53
3 48
4 52
5 50
6 -
Ft+1 = F6 = 51+53+48+52+50/5 = 50.8
50.8

• Is similar to the simple average except that we are not taking average of
all the date, but are including only n of the most recent periods in the
average.
• Where n is a set time period (e.g.: the last four weeks)
• Each new forecast drops the oldest data point & adds a new observation
• More responsive to a trend but still lags behind actual data
• Assumes an average is a good estimator of future behavior
– Used if little or no trend
– Used for smoothing
Simple Moving Average

Problem: Simple Moving Average (1 of 6)
You’re manager in Amazon’s electronics department. You want to forecast ipod
sales for months 4-6 using a 3-period moving average.
Month
Sales
(000)
1 4
2 6
3 5
4 ?
5 ?
6 ?

Month
Sales
(000)
Moving Average
(n=3)
1 4 NA
2 6 NA
3 5 NA
4 ?
5 ?
(4+6+5)/3=5
6 ?
You’re manager in Amazon’s electronics department. You want to forecast ipod
sales for months 4-6 using a 3-period moving average.

What if ipod sales were actually 3 in month 4
Month
Sales
(000)
Moving Average
(n=3)
1 4 NA
2 6 NA
3 5 NA
4 3
5 ?
5
6 ?
?

Forecast for Month 5?
Month
Sales
(000)
Moving Average
(n=3)
1 4 NA
2 6 NA
3 5 NA
4 3
5 ?
5
6 ?
(6+5+3)/3=4.667

Actual Demand for Month 5 = 7
Month
Sales
(000)
Moving Average
(n=3)
1 4 NA
2 6 NA
3 5 NA
4 3
5 7
5
6 ?
4.667?

Forecast for Month 6?
Month
Sales
(000)
Moving Average
(n=3)
1 4 NA
2 6 NA
3 5 NA
4 3
5 7
5
6 ?
4.667
(5+3+7)/3=5

• In simple moving avg. method, each observation is weighted equally. For
example,
– in a three-period moving average each observation is weighted one-third.
– In a five-period moving average each observation is weighted one-fifth.
• Higher or Lower weights are given to observations based on the knowledge of
the industry
• All weights must add to 100% or 1.00,
• From formula given below, e.g. Ct =0.5, Ct-1 = 0.3, Ct-2 =0.2 (weights add to 1.0)
• Allows emphasizing one period over others; above indicates more weight on
recent data (Ct=.5)
• Differs from the simple moving average that weighs all periods equally - more
responsive to trends
Weighted Moving Average

Weighted Moving Average - Problem

• This is a forecasting model that uses sophisticated weighted
average procedure to obtain a forecast.
• Need just three pieces of data to start:
– The current period’s forecast,
– The current period’s actual value
– The value of a smoothing coefficient, α, which varies between 0 and 1.
• The equation for the forecast is quite simple:
– Next period’s forecast = α (current period’s actual) + (1- α) (current
period’s forecast)
• In mathematical terms:
Exponential Smoothing (1 of 4)

• Exponential smoothing models are the most frequently used
forecasting techniques and are available on almost all
computerized forecasting software.
• These models are widely used, particularly in operations
management.
• They have been shown to produce accurate forecasts under
many conditions, yet are relatively easy to use and
understand.

• Problem:

Selecting α:
• Depending on the value of α , you can place more weight on either
the current period’s actual or the current period’s forecast.
• In this manner the forecast can depend more heavily either on
what happened most recently or on the current period’s forecast.
• Values of α that are low— say, 0.1 or 0.2—generate forecasts that
are very stable because the model does not place much weight on
the current period’s actual demand.
• Values of α that are high, such as 0.7 or 0.8, place a lot of weight
on the current period’s actual demand and can be influenced by
random variations in the data.
• Thus, how α is selected is very important in getting a good forecast.

Comparing Forecasts with Different values of α
(1 of 3)
Exponential Smoothing Forecasts
Time Period Actual Demand α = 0.1 α = 0.6
t 0.1 0.6
1 50
2 46 50.00 50.00
3 52 49.60 47.60
4 51 49.84 50.24
5 48 49.96 50.70
6 45 49.76 49.08
7 52 49.28 46.63
8 46 49.56 49.85
9 51 49.20 47.54
10 48 49.38 49.62

(2 of 3)
• NOTE:
• When using the exponential smoothing equation, always make sure you have the three pieces of
information needed:
1. The current period’s forecast,
2. The current period’s actual value, and
3. A value for the smoothing coefficient, α.
This problem illustrates how you can begin the exponential smoothing process when you do not have
initial forecast values.
• From table above, notice that we used the naïve method to derive initial values of forecasts for
period 2.
• Then to obtain forecasts for period 3, we used the exponential smoothing equation with
different values of α.
For an α = 0.10, the forecast for period 3 is computed as:
F3 = (0.10)(46)+(0.90)(50) = 49.6
For an α = 0.60, the forecast for period 3 is computed as:
F3 = (0.60)(46)+(0.40)(50) = 47.6

(3 of 3)
40
42
44
46
48
50
52
54
1 2 3 4 5 6 7 8 9 10
Actual Demand
α = 0.1
α = 0.6

Forecasting: Causal Models

• Often, leading indicators can help to predict changes in future
demand e.g. housing starts
• Causal models establish a cause-and-effect relationship
between independent and dependent variables
• A common tool of causal modeling is linear regression:
• Additional related variables may require multiple regression
modeling
Causal Models
bxaY 

Author: Niranjana K R 57
Linear Regression
  
   
 



XXX
YXXY
b
2
• Identify dependent (y) and
independent (x) variables
• Solve for the slope of the line
• Solve for the y intercept
• Develop your equation for the
trend line
Y= a + bX
XbYa 




 22
XnX
YXnXY
b
24-08-2016

Linear Regression Problem: A maker of golf shirts has been tracking the relationship
between sales and advertising dollars. Use linear regression to find out what sales
might be if the company invested $53,000 in advertising next year.




 22
XnX
YXnXY
bSales $
(Y)
Adv.$
(X)
XY X^2 Y^2
1 130 32 4160 2304 16,900
2 151 52 7852 2704 22,801
3 150 50 7500 2500 22,500
4 158 55 8690 3025 24964
5 153.85 53
Tot 589 189 28202 9253 87165
Avg 147.25 47.25
  
 
 
  153.85531.1592.9Y
1.15X92.9bXaY
92.9a
47.251.15147.25XbYa
1.15
47.2549253
147.2547.25428202
b 2








24-08-2016

Correlation Coefficient How Good is the Fit?
• Correlation coefficient (r) measures the direction and strength of the linear
relationship between two variables. The closer the r value is to 1.0 the better the
regression line fits the data points.
• Coefficient of determination ( ) measures the amount of variation in the dependent
variable about its mean that is explained by the regression line. Values of ( ) close to
1.0 are desirable.
    
       
   
   
  .964.982r
.982
58987,1654*(189)-4(9253)
58918928,2024
r
YYn*XXn
YXXYn
r
22
22
2
2
2
2










2
r
2
r
24-08-2016

Multiple Regression
• An extension of linear regression but:
– Multiple regression develops a relationship between a dependent
variable and multiple independent variables. The general formula is:
24-08-2016

Forecast Errors
• Error is the difference between the forecast value and what
actually occurred.
• In statistics, these are called residuals.
• As long as the forecast value is within the confidence limits,
this is really not an error.
• Demand for a product is generated through the interaction of
a number of factors too complex to describe accurately in a
model.
• So, every forecasts contain some error.
• We must differentiate, sources of error and the measurement
error.

Sources of Errors
• Errors can come from variety of sources
• Projecting past trends into the future is one of the common source
– Example: When we talk about statistical errors in regression analysis, we are
referring to the deviations of observations from our regression line.
• Errors can be classified as:
– Bias Errors
• Occur when a consistent mistake is made,
• Failure to include the right variables is a source,
• Wrong relationships among variables,
• Employing wrong trend line,
• A mistaken shift in the seasonal demand from where it normally occurs, and
• The existence of some undetected secular trend.
– Random Errors
• These errors can be defined as those that cannot be explained by the forecast
model being used.

Measurement of Error
Forecast errors are defined as:
et = Forecast error = Actual demand for period t – Forecast for period t = At - Ft
• Forecast errors provide a measure of accuracy and a basis for comparing the
performance of alternate models. Commonly used error measures are:

Selecting the Right Forecasting Model
1. The amount & type of available data
 Some methods require more data than others
2. Degree of accuracy required
 Increasing accuracy means more data
3. Length of forecast horizon
 Different models for 3 month vs. 10 years
4. Presence of data patterns
 Lagging will occur when a forecasting model meant for a
level pattern is applied with a trend

Forecasting Software
• Spreadsheets
– Microsoft Excel, Quattro Pro, Lotus 1-2-3
– Limited statistical analysis of forecast data
• Statistical packages
– SPSS, SAS, NCSS, Minitab
– Forecasting plus statistical and graphics
• Specialty forecasting packages
– Forecast Master, Forecast Pro, Autobox, SCA
24-08-2016

Mba om 05_forecasting_foroperations

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to Mba om 05_forecasting_foroperations

Similar to Mba om 05_forecasting_foroperations (20)

More from Niranjana K.R.

More from Niranjana K.R. (13)

Recently uploaded

Recently uploaded (20)

Mba om 05_forecasting_foroperations