Professional Documents
Culture Documents
Chapter 3 - Forecasting PDF
Chapter 3 - Forecasting PDF
CSSBB
❖ EXPECTED LEVEL OF
DEMAND
❖ The level of demand may be
a function of some structural
variation such as trend or
seasonal variation
❖ ACCURACY
❖ potential size of forecast
error
FORECAST USES
1. PLAN THE SYSTEM
2. PLAN THE USE OF THE
SYSTEM
FORECAST FEATURES
❖ If it happened in the past, it
will happen in the future
❖ Not perfect
❖ Random Variation = residual
error
❖ Group forecast is more
accurate than individual
forecast
❖ Accuracy decreases overtime
ELEMENTS OF A GOOD FORECAST
❖ TIMELY
❖ ACCURATE
❖ RELIABLE
❖ EXPRESSED IN
MEANINGFUL UNITS
❖ IN WRITING
❖ SIMPLE TO UNDERSTAND
❖ COST-EFFECTIVE
STEPS IN THE FORECASTING PROCESS
❖ ALLOWANCES SHOULD BE
MADE FOR FORECAST
ERRORS
❖ EX. MARGIN OF ERROR
❖ FORECAST ERRORS
SHOULD BE MONITORED
❖ ERROR/RESIDUAL =
ACTUAL DATA -
FORECASTED DATA
FORECASTING ACCURACY METRICS
❖ MEAN ABSOLUTE
DEVIATION (MAD)
❖ MEAN SQUARED ERROR
(MSE)
❖ MEAN ABSOLUTE
PERECENT ERROR (MAPE)
FORECAST ACCURACY METRICS
∑ Actual t − Forecast t MAD weights all errors
MAD =
n evenly
2
∑ (Actual t − Forecast t ) MSE weights errors according
MSE = to their squared values
n −1
Actualt − Forecast t
∑ × 100
Actualt MAPE weights errors
MAPE = according to relative error
n
LO 3.5
FORECAST ERROR CALUCLATION
Actual Forecast (A-F)
Period
(A) (F) Error |Error| Error2 [|Error|/Actual]x100
1 107 110 -3 3 9 2.80%
Sum 13 39 11.23%
∑A
i =1
t −i
At − n + ... + At − 2 + At −1
Ft = MA n = =
n n
where
Ft = Forecast for time period t
MA n = n period moving average
At −i = Actual value in period t − i
n = Number of periods in the moving average
QUANTITATIVE FORECASTING
❖ Using a three-point moving PERIOD DEMAND
average, whats the forecasted
demand for period 6? 1 42
❖ F6 = 43 + 40 + 31 / 3 = 41.33
2 40
Ft = wt ( At ) + wt −1 ( At −1 ) + ... + wt − n ( At −n )
where
wt = weight for period t , wt −1 = weight for period t − 1, etc.
At = the actual value for period t , At −1 = the actual value for period t − 1, etc.
QUANTITATIVE FORECASTING
❖ Compute a weighted forecasting
PERIOD DEMAND
using a weight of 40% for the most
recent and 30% for the next most
recent, 20% for the next and 10% for 1 42
the next
❖ F6 = .10(40) + .20(43) + .30(40) + . 2 40
40(41) = 41.0
3 43
❖ If the actual demand for period 6 is
39, forecast demand for period 7 using
4 40
the same weights
❖ F7 = .10(43) + .20(40) + .30(41) + .
40(39) = 40.2 5 41
QUANTITATIVE FORECASTING
❖ EXPONENTIAL SMOOTHING
❖ A weighted averaging method that is based on the previous
forecast plus a percentage of the forecast error.
Ft = Ft −1 + α ( At −1 − Ft −1 )
where
Ft = Forecast for period t
Ft −1 = Forecast for the previous period
α = Smoothing constant
At −1 = Actual demand or sales from the previous period
QUANTITATIVE FORECASTING
Actual
❖ The symbol alpha represents a Period (t)
Demand
Forecast 1 Error Forecast 2 Error
forecast error
4 40 41.92 -1.92 41.92 -1.92
is 43, so:
7 46 41.39 4.61 40.25 5.75
Ft = a + bt
where
Ft = Forecast for period t
a = Value of Ft at t = 0
b = Slope of the line
t = Specified number of time periods from t = 0
TECHNIQUES FOR TREND
❖ Plot the data Week (x) Unit Sales (y) xy
1 700 700
❖ Does the plot suggests that
2 724 1448
a linear trend line would be
3 720 2160
appropriate?
4 728 2912
❖ If yes, compute the 5 740 3700
coefficients of the trend line 6 742 4452
7 758 5306
8 750 6000
9 770 6930
10 775 7750
TOTAL 7407 41358
TECHNIQUES FOR SEASONALITY
❖ TWO MODELS OF
SEASONALITY:
❖ MULTIPLICATIVE
❖ SEASONAL RELATIVES
TECHNIQUES FOR SEASONALITY
❖ ADDITIVE
❖ Seasonality is expressed as a
quantity that gets added to or
subtracted from the time-series
average in order to incorporate
seasonality
❖ MULTIPLICATIVE
❖ Seasonality is expressed as a
percentage of the average
amount which is then used to
multiply the value of a series in
order to incorporate seasonality
TECHNIQUES FOR SEASONALITY
A furniture manufacturer Period Quarter Sales Quarter Deseasonalized
wants to predict Rela0ve Sales
quarterly demand for a 1 1 132 1.20 110
certain furniture for 2 2 140 1.10 127.27
periods 15 and 16 3 3 146 0.75 194.67
which happens to be 4 4 153 0.95 161.05
the second and third
quarters of a particular 5 1 160 1.20 133.33
y e a r. T h e s e r i e s 6 2 168 1.10 152.73
composed of both trend 7 3 176 0.75 234.67
and seasonality.
8 4 185 0.95 194.74
The trend portion of
demand is projected
using the equation Ft =
124 + 7.5t.
Quarter relatives are
Q1= 1.20 Q2=1.10
Q3=0.75 and Q4 = 0.95
TECHNIQUES FOR SEASONALITY
Use the information to predict demand for periods
15 and 16.
y = a + bx
where:
y = predicted (dependent) variable
x = predicted (independent)
variable
b = slope of the line
a = intercept
SIMPLE LINEAR REGRESSION
❖ The coefficient a and b of the
line are boased on the
following two equations:
SIMPLE LINEAR REGRESSION
❖ Plot the data and decide if a UNIT SALES, x PROFITS, y
linear model is reasonable 7 $0.15
2 0.10
❖ Using excel, obtain regression 6 0.13
equation 4 0.15
14 0.25
❖ y = 0.0506 + 0.0159x
15 0.27
❖ If predicted sales is 10, then 16 0.24
profits would be .2099 or 12 0.20
$209,900. 14 0.27
20 0.44
15 0.34
7 0.17
CORRELATION
❖ Measure the strenght the
direction of relationship
between two variables.
❖ can rage from -1.00 to +1.00
❖ +1.00 indicates a positive
correlation, -1.00 indicates a
negative correlation
CORRELATION
❖ Correlation is measured by R
• COST
• ACCURACY
• AVAILABILITY OF
HISTORICAL DATA
• AVAILABILITY OF
FORECASTING SOFTWARE
• TIME NEEDED TO GATHER
AND ANALYZE DATA AND
PREPARE FORECAST
• FORECAST HORIZON
SEATWORK
❖ GET A ONE WHOLE SHEET OF YELLOW PAD PAPER
❖ ANSWER THE CASE STUDY IN EDMODO