Midterm- Forecasting Flashcards
statement about the future value of a variable of interest
forecast
forecast of day to day operations
short range forecasts
(S/L)types of products and services to offer
L
(S/L) Facilitites and equipments to have
L
(S/L) Location
L
(S/L) Scheduling
S
planning inventory and work force levels(S/L)
S
(S/L) purchasing and budgeting
S
involves ling range plans about the types of products and services to offer
Plan the system
short range and intermediate-range planning
Plan the use of the system
2 Uses of forecast
plan the system
plan the use
new product/process cost estimates (Bus. Org)
Accounting
(Bus. Org) equipment replacement needs
finance
hiring activities (Bus. Org)
HR
pricing and promotion (Bus. Org)
Marketing
new/revised info system (Bus. Org)
MIS
work assignments and work loads (Bus. Org)
Ops
revision of current features (Bus. Org)
product and service design
process cost estimates (Bus. Org)
acounting
outsourcing (Bus. Org)
operations
internet services (Bus. Org)
mis
e-business strategies (Bus. Org)
marketing
layoff planning (Bus. Org)
HR
cash management (Bus. Org)
accounting
funding/borrowing (Bus. Org)
finance
profit projections (Bus. Org)
accounting
outplacement (Bus. Org)
HR
global comp strategies (Bus. Org)
marketing
porject management (Bus. Org)
operations
Features common to all forecasts (4)
causal system that existed in the past will continue to exist in the future
rarely perfect
group items more accurate
accuracy decreases as time horizon increases
allowances should be made for____
errors
forecasting errors among items in a group have a ______
cancelling effect
grouping may arise if ____ are used to make _____products
raw materials
multiple
time period covered by forecast
time horizon
Elements of a good forecast (7)
timely accurate reliable meaningful units in writing simple to understand and use cost-effective
enabling users to plan against possible errors&provide basis for comparing alternative forecast
accurate
forecast should work consistently
reliable
it will increase likelihood tgat all concerened are using the same information
in writing
benefits should outweigh the cost
cost effective
______ will permit an objective basis for evaluating the forecast once actual results are in
written forecast
6 Basic steps in forecasting process
- determine the purpose of the forecast
- establish time horizon
- select a forecasting technique
- obtain, clean and analyze appropriate data
- make the forecast
- monitor the forecest
2 general approaches to forecasting
qualitative
quantitative
subjective inputs&soft information
qualitative
objective inputs&hard data
quantitative
defies numerical description
qualitative
defies personal biases
quantitative
forecast that use subjective inputs
judgmental
project patterns identified in recent time-series observations
time series forecast
uses explanatory variables to predict future demand
associative model
Judgmental forecast (4)
executive
sales force
consumer
delphi approach
upper level managers (finance, matketing, manufacturing managers) join to prepare forecast
executive opinions
joint estimates of sales people and customer service people
sales force opinion
managers and staff complete a series of questionnaires
delphi approach
good sources of info because of their ditect contact with consumers
sales staff and customer service staff
determines demand
consumers
time-ordered sequence of observations taken at regular intervals
time-series
analysis of time-series data requires
identi of underlying behavior of the series
identification kf underlying behavior of series is done by
plotting the data
long-term upward or downward movement in dat
trend
short term regular variations to the calendar or time of the day
seasonality
wavelike variations lasting more than one year
cycles
caused by unusual circumstances and are not reflective if typical behavior
irregular variation
residual variations that remain after all other behaviors have been accounted for
random variation
populations shifts, changing incomes, cultural changes
trend
economic, political and agricultural conditions
cycles
may be caused by severe weather conditons, strikes, major change in product or service
irregular variation
forecast for any period that equals the previous period actual value
naive forecast
basis of forecast
single previous value of a time series
last data point=forecast for the next period
naive used with stable series
forecast this season=value of the series of last season
naive used with seasonal variations
last value of the series +|- the difference between the last two values of the series
naive used with trend
Advantage if naive (3)
no cost
quick and easy to prep
data analysis is non existent thats why easily understandable
Disadvantage of naive
inability to provide highly accurate forecast
smooth variation in data
averaging techniques
small variations
random
large variations
real variations
reflect recent value of time series
averaging tech
average value ocer the last several periods
averaging tech
3 techniques for averaging
moving
weighted moving
exponential smoothing
averages a number of recent actual values, updated as new values become available
moving average
more recent values in a series are given more weight in computing a forecast
weighted moving
weighted moving average based on previous forecast plus a percentage of the forecast error
exponential smoothing
previous forecast plus the difference with such forecast and the actual value of the series at that point
exponential smoothing
next value in a series will ewual the previous value in comparable period
naive
forecast is based in an average of recent values
moving average
sophisticated form of weighted moving
exponential smoothing
2 important techniques to develop forecast when trend is present
1 trend equation
2 trend-adjusted exponential smoothing
used to develop forecast when trend is present
linear trend equation
variation of exponential smoothing used when a time series exhibits linear trend
trend adjusted exponential smoothing
2 elements of trend adjusted forecast
smoothed error
trend factor
a forecast model for trend
adjusted expo smoothing
regularly repeating movements in series values that can be tied up to recurring events
seasonal variations
may refer to regular annual variations
seasonality
percentage of average ir trend
seasonal relatives
2 models of seasonality
additive
multiplicative
seasonalityis expressed as a quantity
additive model
seasonality is expressed as a percentage of a trend
multiplicative model
two uses of seasonality
deseasonalize data
incorporate seasonality in a forecast
removing the seasonal components from data in order to get a picture of non seasonal components
deseasonalize data
useful when demand has both trend and seasonal component
incorporate seasonality in a forecast
dividing each data point by its corresponding seasonal relative
deseasonalize
obtaining trend estimates using trend equation.
add seasonality to the trend estimates by multiplying these trend estimates by corresponding seasonal relatives
Incorporte seasonality in forecast
up and down movements similar to seasonal variations but of longer duration (2-6 yrs)
cycles
search doe another variable that relates to and leads the variable of interest
cycles
Time series forecasts (6)
Naive averaging trend trend adjusted expo smoothing seasonality cycles
Associative forecasting techniques (3)
simple linear regression
comments on the use of linear regression analysis
curvilinear and multiple regression analysis
rely on identification of related variables that can be used to predict bvalues of the variable of interest
associative forecasting tech
associative tech has an equation that summarizes the effects of____
predictor variables
the primary method used of analysis
regression
it is a technique for fitting a line to a set of points
regression
simplest and widely used form of regression
simple linear regression
involves a linear relationship bet. two variables
simple linear regression
minimizes the sum of the squared vertical deviations around the line
least square line
uncontrollable bariables that tend to lead or precede changes on a variable of interest
indicators
3 conditions for an indicator to be valid
- indi and varia should have logical explanations
- indicator must precede dependent variables; forecase isnt outdated
- small corellation may imply that other variables are important
weaknesses of regression (3)
Applies obly to linear relationships with one independent variable
needs considerable amount of data
all observations are weighted equally
measure the strength and direction of relationship bet. 2 variables
correlation
Comments on the use of linear regression anaylsis (3)
variations around the line are random
deviations around the line be normally distributed
predictions within the range
_____ the data to verify that a linear relationship is appropriate
always plot
_____may be time dependent
data
_____may imply that other variables are important
small correlation
when non linear relationship are present
curvilinear regression
modles that innvolve more than one predictor
multiple regression analysis
______substantially increases data requirments
multiple data analysis
basis of orgs schedules
forecasts
difference between the actual value and the value that was predicted for a given period
forecast error
actual-forecast
error
significant factor to decide among forecasting alternatives
forecast accuracy
average absolute forecast error
mean absolute deviation
average of swuared forecast errors
mean squared error
the average absolute percent error
mean absolute percent error
will provide insight in WON forecasts are performing satisfactorily
tracking and analysis of forecast errors
forecast is deemed to perform adequately if errors show only____
random variations
inherent variation, remains in data, even after all causes for variation has been accounted for
random variations
cisual toll for monitoring forecast error
control chart
center line means
zero error
How to construct control chart (3)
compute MSE
compute for the upper control limit
lower limit
the ratio of cumulative forecast error to corresponding value of MAD, used to forecast
tracking signal
its purpose is to detect ant bias in errors
tracking signal
tendency for sequence of errors to be postive or negative
bias
values outside of limits means
there is bias in forecast
two most important factors in choosing forecasting tech
cost
accuracy
SHORT prep time
movigg average
simple expo
trend adjusted
trend models
Short-moderate prep time
seasonal
long develpment prep time
causal regression models
Stationary data pattern (2)
moving ave
simple expo
Trend data pattern
adjusted
trend models
complex patterns
causal regression models
Short forecast horizon
moving ave
simple expo
short to medium forcast horizon
trend
seasonal
short medium long forecast horizon
causal regre
2-3 observations
moving average
5-10 observations
simple expo
10-15 observations
trend adjused
10-20, 5 pee season if seasonal
trend models
2 peaks and troughs
seasonal
10 obs per independent variable
causal reg
2 approaches to forecat
reactive
proactive
views forecast as probable future demand
reactive
(approach)
adjust production rates
reactive
inventories (approach)
reactive
workforce (approach)
reactive
seeks to influence the demand
proactive
advertising (approach)
proactive
pricing, product changes (approach)
proactive
