Scribd
Upload
130 views4 pages

Chapter 18 Operations Management

1. Forecasting involves using past data to predict future values and guide management decisions. There are qualitative, time series, causal, and simulation forecasting methods. 2. Time series analysis uses past demand data in simple or weighted moving averages, exponential smoothing, or regression models to forecast short, medium, and long term. 3. Exponential smoothing is commonly used as it is accurate, easy to use, and requires little computation or data storage. It uses a smoothing constant to calculate forecasts based on past forecasts and actual data.

Uploaded by

Greta Zanatta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
130 views4 pages

Chapter 18 Operations Management

1. Forecasting involves using past data to predict future values and guide management decisions. There are qualitative, time series, causal, and simulation forecasting methods. 2. Time series analysis uses past demand data in simple or weighted moving averages, exponential smoothing, or regression models to forecast short, medium, and long term. 3. Exponential smoothing is commonly used as it is accurate, easy to use, and requires little computation or data storage. It uses a smoothing constant to calculate forecasts based on past forecasts and actual data.

Uploaded by

Greta Zanatta
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

FORECASTING

= the average of a fixed number of past periods


• Affects significant management decisions
- Finance & accounting: budgeting and cost control
- Marketing: product planning and personnel compensation
- Selection of suppliers, capacity requirements and decisions about purchasing,
staffing and inventory

Strategic forecasts
• Guide day-to-day decisions

Decoupling point = point at which inventory is stored (supply chain operate independently)
• Strategic forecasting determines level of inventory needed at decoupling points

4 types of forecasts
1. Qualitative
2. Time series analysis: forecasting basing on past demand
3. Casual relationships
4. Simulation

Cyclical factors are more difficult to determine (caused by chance events: war, economic
conditions, natural disasters, etc.)
• Autocorrelation / serial correlation: the value expected at any point is highly correlated
with its own past values  past values of the variables can be used to predict future
values

Identification of the trend lines is a starting point when developing a forecast


I. Linear
II. S-curve
III. Asymptotic
IV. Exponential

Qualitative forecasting
Used to take advantage of expert knowledge
• Useful when judgment is required, products are new, or the firm has little experience in
a new market (market research)

Time series analysis


I. Short term (x > 3 months): tactical decisions
1. Simple moving average
2. Weighted moving average and simple exponential smoothing => 5 to 10
observations (stationary data)
3. Exponential smoothing with trend => 5 to 10 observations (stationary and trend
data)
II. Medium term (3 months < x < 2 years): seasonal effects
4. Linear regression => 10 to 20 observations (stationary, trend and seasonality data)
5. Trend and seasonal models => 2 to 3 observations per season (stationary, trend
and seasonality data)
III. Long term (2 years < x): general trends and major turning points

Factors affecting forecasting model selection


• Time horizon to be forecasted
• Data availability
• Accuracy required
• Size of forecasting budget
• Availability of qualified personnel

Simple moving average


• 6 to 12 months
• Weekly data
• Stationary data pattern
Useful: demand is not growing or declining rapidly and so seasonality is present
• Removes some random fluctuations from the data
• Selecting the period length is important
- Longer periods => more smoothing
- Shorter periods => react to trend more quickly
• The formula implies equal weighting for all periods:
A + A + A +...+ A t−n
F t= t −1 t−2 t −3
n

Weighted moving average


• 6 to 10 observations needed to start
• Stationary data pattern
• The formula allows unequal weighting of prior time periods
- The sum of the weights must be equal to 1
• Most recent periods have higher weights than farther periods in the past (experience
and trial-and-error are the simplest approaches)
At −1+ A t−2 + At −3+...+ A t−n
F t=
n

Exponential smoothing
= a weighted average method that includes all past data
• An integral part of computerized forecasting
• The most used forecasting technique
- Accurate
- Easy
- User-understandable
- Little computation required
- Small computer storage
- Easy-to-compute accuracy tests
• Only 3 pieces of data required
1. Most recent forecast
2. Actual demand for the forecast period
3. Smoothing constant alpha (desired response / reaction rate)
F t=F t −1 +α ( A t −1−F t −1)

Exponential smoothing with trend


• 6 to 10 observations needed to start
• Stationary and trend data pattern
Sometimes the exponential smoothing fails to keep up with the actual data => can be
corrected by adding a trend adjustment using two constants
• Smoothing constant alpha
• Trend smoothing constant delta
The constants have a given value between 0 and 1: usually small values are used (0.1 to
0.3, depending on how much random variation there is in demand)

Linear regression analysis


• Useful for long-term forecasting
• 10 to 20 observations needed to start
• Stationary, trend and seasonality data pattern
Regression is used to identify the relationship between 2 or more correlated variables (from
observed data)
• The dependent variable is predicted for given values of the independent variable
• The relationship between the variables is assumed to be explained with a straight line

Time series decomposition


Time series = chronologically ordered data
• May contain: trend, seasonal, cyclical, autocorrelation and random
Time series decomposition = separating the time series data into these components

Seasonal factors = past sales / average sale for each season

Example:
In the past years, a firm sold an average of 1000 units each year
• Spring: 200
• Summer: 350
• Fall: 300
• Winter: 150

Average sale for each season: 1000/4 = 250


Seasonal factors:
• Spring: 200/250 = 0.8
• Summer: 350/250 = 1.4
• Fall: 300/250 = 1.2
• Winter: 150/250 = 0.6
The company expects demand for next year to be 1100 units
Demand per period:
• Spring => 1000 : 200 = 1100 : x => 220
• Summer => 1000 : 350 = 1100 : x => 380
• Fall => 1000 : 300 = 1100 : x => 330
• Winter => 1000 : 150 = 1100 : x => 165

Forecast errors
= the difference between the forecast value and what actually occurred
• All forecasts contain some level of error
• Sources of error:
- Bias => when a consistent mistake is made
- Random => errors that are not explained by the model being used

Measures of error
• Mean absolute deviation (MAD)
- Ideally, it will be 0: larger values indicate less accuracy
• Mean absolute percent error (MAPE)
- Scales the forecast error to the magnitude of demand
• Tracking signal
- Indicates whether forecast errors are accumulating over time (either positive or
negative errors)

Causal relationship forecasting


Uses independent variables other than time to predict future demand
• This independent variable must be a leading indicator (using rain to forecast sales of
umbrellas)
• Many apparently causal relationships are actually just correlated events

Multiple regression techniques


• Often, more than 1 independent variable may be a valid predictor of future demand
- Forecast analyst uses multiple regression (= like linear regression, but with multiple
independent variables)

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy