Forecasting

Forecasting Models
Forecasting
Techniques

Qualitative Time Series

Models Methods

Delphi
Naive
Method
Moving
Jury of Executive
Average
Opinion
Weighted
Sales Force
Moving Average
Composite
Exponential
Consumer Market
Smoothing
Survey

Trend Analysis
Causal
Methods
Seasonality
Simple Analysis
Regression
Analysis Multiplicative
Decomposition
Multiple
Regression
Analysis
Model Differences
 Qualitative – incorporates judgmental &
subjective factors into forecast.
 Time-Series – attempts to predict the
future by using historical data.
 Causal – incorporates factors that may
influence the quantity being forecasted
into the model
Qualitative Forecasting Models
 Delphi method
 Iterativegroup process allows experts to make
forecasts
 Participants:
 decision makers: 5 -10 experts who make the forecast
 staff personnel: assist by preparing, distributing, collecting,
and summarizing a series of questionnaires and survey
results
 respondents: group with valued judgments who provide input
to decision makers
Qualitative Forecasting Models (cont)
 Jury of executive opinion
 Opinions of a small group of high level managers, often in
combination with statistical models.
 Result is a group estimate.
 Sales force composite
 Each salesperson estimates sales in his region.
 Forecasts are reviewed to ensure realistic.
 Combined at higher levels to reach an overall forecast.
 Consumer market survey.
 Solicits input from customers and potential customers regarding
future purchases.
 Used for forecasts and product design & planning
Forecast Error Forecast Error  At  Ft

 Bias - The arithmetic sum of T

the errors MSE  

t 1
| forecast error | 2
/T
 Mean Square Error - Similar to T
simple sample variance   (At  Ft ) 2 / T
 Variance - Sample variance t 1
(adjusted for degrees of
freedom)
 Standard Error - Standard
deviation of the sampling
distribution
T T
 MAD - Mean Absolute MAD   | forecast error | /T   |At  Ft | / T
Deviation t 1 t 1
 MAPE – Mean Absolute
Percentage Error T
MAPE  100 [|At  Ft | / At ] / T
t 1
 Quantitative Forecasting Models
 Time Series Method
 Naïve
 Whatever happened
recently will happen
again this time (same
Ft  Yt 1
time period)
 The model is simple

Ft  Yt  4 : Quarterly data
and flexible
 Provides a baseline to
measure other models
Ft  Yt 12 : Monthly data
 Attempts to capture
seasonal factors at the
expense of ignoring
trend
Naïve Forecast
Wallace Garden Supply
Forecasting
Storage Shed Sales

Actual Naïve Absolute Percent Squared

Period Value Forecast Error Error Error Error
January 10 N/A
February 12 10 2 2 16.67% 4.0
March 16 12 4 4 25.00% 16.0
April 13 16 -3 3 23.08% 9.0
May 17 13 4 4 23.53% 16.0
June 19 17 2 2 10.53% 4.0
July 15 19 -4 4 26.67% 16.0
August 20 15 5 5 25.00% 25.0
September 22 20 2 2 9.09% 4.0
October 19 22 -3 3 15.79% 9.0
November 21 19 2 2 9.52% 4.0
December 19 21 -2 2 10.53% 4.0
9.000 3 17.76% 10.091
BIAS MAD MAPE MSE

Standard Error (Square Root of MSE) = 3.176619

Naïve Forecast Graph
Wallace Garden - Naive Forecast

25

20

15
Sheds

Actual Value
Naïve Forecast
10

0
February March April May June July August September October November December
Period
 Quantitative Forecasting Models
 Time Series Method
 Moving Averages
 Assumes item
forecasted will stay
steady over time.
 Technique will smooth
out short-term
irregularities in the time
series.
k
k - period moving average   (Actual value in previous k periods) /k
k 1
Moving Averages
Wallace Garden Supply
Forecasting
Storage Shed Sales

Actual
Period Value Three-Month Moving Averages
January 10
February 12
March 16
April 13 10 + 12 + 16 / 3 = 12.67
May 17 12 + 16 + 13 / 3 = 13.67
June 19 16 + 13 + 17 / 3 = 15.33
July 15 13 + 17 + 19 / 3 = 16.33
August 20 17 + 19 + 15 / 3 = 17.00
September 22 19 + 15 + 20 / 3 = 18.00
October 19 15 + 20 + 22 / 3 = 19.00
November 21 20 + 22 + 19 / 3 = 20.33
December 19 22 + 19 + 21 / 3 = 20.67
Moving Averages Forecast
Wallace Garden Supply
Forecasting 3 period moving average
Actual Value - Forecast

Input Data Forecast Error Analysis

Absolute Squared Absolute
Period Actual Value Forecast Error error error % error
Month 1 10
Month 2 12
Month 3 16
Month 4 13 12.667 0.333 0.333 0.111 2.56%
Month 5 17 13.667 3.333 3.333 11.111 19.61%
Month 6 19 15.333 3.667 3.667 13.444 19.30%
Month 7 15 16.333 -1.333 1.333 1.778 8.89%
Month 8 20 17.000 3.000 3.000 9.000 15.00%
Month 9 22 18.000 4.000 4.000 16.000 18.18%
Month 10 19 19.000 0.000 0.000 0.000 0.00%
Month 11 21 20.333 0.667 0.667 0.444 3.17%
Month 12 19 20.667 -1.667 1.667 2.778 8.77%
Average 12.000 2.000 6.074 10.61%
Next period 19.667 BIAS MAD MSE MAPE
Moving Averages Graph
Three Period Moving Average

25

20

15
Value

Actual Value
Forecast

10

0
1 2 3 4 5 6 7 8 9 10 11 12
Time
 Quantitative Forecasting Models
 Time Series Method
 Weighted Moving Averages
 Assumes data from some periods are more
important than data from other periods (e.g.
earlier periods).
 Use weights to place more emphasis on some
periods and less on others.

k - period weighted moving average 

k k

 (Weight for each period i)(Actual value in previous k periods) /  (weights)

i 1 i 1
Weighted Moving Average
Wallace Garden Supply
Forecasting
Storage Shed Sales

Actual
Period Value Weights Three-Month Weighted Moving Averages
January 10 0.222
February 12 0.593
March 16 0.185
April 13 2.2 + 7.1 + 3 / 1 = 12.298
May 17 2.7 + 9.5 + 2.4 / 1 = 14.556
June 19 3.5 + 7.7 + 3.2 / 1 = 14.407
July 15 2.9 + 10 + 3.5 / 1 = 16.484
August 20 3.8 + 11 + 2.8 / 1 = 17.814
September 22 4.2 + 8.9 + 3.7 / 1 = 16.815
October 19 3.3 + 12 + 4.1 / 1 = 19.262
November 21 4.4 + 13 + 3.5 / 1 = 21.000
December 19 4.9 + 11 + 3.9 / 1 = 20.036

Next period 20.185

Sum of weights = 1.000

 Weighted Moving Average
Wallace Garden Supply
Forecasting 3 period weighted moving average

Input Data Forecast Error Analysis

Absolute Squared Absolute %
Period Actual value Weights Forecast Error error error error
January 10 0.222
February 12 0.593
March 16 0.185
April 13 12.298 0.702 0.702 0.492 5.40%
May 17 14.556 2.444 2.444 5.971 14.37%
June 19 14.407 4.593 4.593 21.093 24.17%
July 15 16.484 -1.484 1.484 2.202 9.89%
August 20 17.814 2.186 2.186 4.776 10.93%
September 22 16.815 5.185 5.185 26.889 23.57%
October 19 19.262 -0.262 0.262 0.069 1.38%
November 21 21.000 0.000 0.000 0.000 0.00%
December 19 20.036 -1.036 1.036 1.074 5.45%
Average 12.326 1.988 6.952 10.57%
Next period 20.185 BIAS MAD MSE MAPE

Sum of weights = 1.000

Quantitative Forecasting Models
 Time Series Method
 Exponential Smoothing
 Moving average technique that requires little
record keeping of past data.
 Uses a smoothing constant α with a value between
0 and 1. (Usual range 0.1 to 0.3)

Forecast for period t 

forecast for period t - 1   (actual value in period t - 1 - forecast for period t - 1)
Exponential Smoothing Data
Storage Shed Sales

Exponential Smoothing
Actual
Period Value Ft α At Ft Ft+1
January 10 10 0.1
February 12 10 + 0.1 *( 10 - 10 ) = 10.000
March 16 10 + 0.1 *( 12 - 10 ) = 10.200
April 13 10 + 0.1 *( 16 - 10 ) = 10.780
May 17 11 + 0.1 *( 13 - 11 ) = 11.002
June 19 11 + 0.1 *( 17 - 11 ) = 11.602
July 15 12 + 0.1 *( 19 - 12 ) = 12.342
August 20 12 + 0.1 *( 15 - 12 ) = 12.607
September 22 13 + 0.1 *( 20 - 13 ) = 13.347
October 19 13 + 0.1 *( 22 - 13 ) = 14.212
November 21 14 + 0.1 *( 19 - 14 ) = 14.691
December 19 15 + 0.1 *( 21 - 15 ) = 15.322
Exponential Smoothing
Forecasting Exponential smoothing

Input Data Forecast Error Analysis

Absolute Squared Absolute
Period Actual value Forecast Error error error % error
Month 1 10 10.000
Month 2 12 10.000 2.000 2.000 4.000 16.67%
Month 3 16 10.200 5.800 5.800 33.640 36.25%
Month 4 13 10.780 2.220 2.220 4.928 17.08%
Month 5 17 11.002 5.998 5.998 35.976 35.28%
Month 6 19 11.602 7.398 7.398 54.733 38.94%
Month 7 15 12.342 2.658 2.658 7.067 17.72%
Month 8 20 12.607 7.393 7.393 54.650 36.96%
Month 9 22 13.347 8.653 8.653 74.879 39.33%
Month 10 19 14.212 4.788 4.788 22.925 25.20%
Month 11 21 14.691 6.309 6.309 39.806 30.04%
Month 12 19 15.322 3.678 3.678 13.529 19.36%
Average 5.172 31.467 28%
Alpha 0.100 MAD MSE MAPE

Next period 15.690

Exponential Smoothing
Exponential Smoothing

25

20

15
Sheds

Actual value
Forecast

10

0
y ri l ly st r r r r
ar ar
y c h ay ne Ju be be be be
nu ru ar Ap M Ju gu em cto em em
a b M Au t v c
J Fe Se
p O
No De
 Reqd:
Seatwork
Period Actual Value Compute for the
1 31 value of
2 28 BIAS,MAD,
3 30 MAPE, MSE using
4 27 3 Month Weighted
5 26 Moving Average
6 28
7 31 Weights:
8 29 1st- 0.40
9 33 2nd -0.30
10 28 3rd -0.30
Trend & Seasonality
 Trend analysis
 technique that fits a trend equation (or curve) to a series of
historical data points.
 projects the curve into the future for medium and long term
forecasts.
 Seasonality analysis
 adjustment to time series data due to variations at certain
periods.
 adjust with seasonal index – ratio of average value of the item in
a season to the overall annual average value.
 example: demand for coal & fuel oil in winter months.
Linear Trend Analysis
Midwestern Manufacturing Sales

Sales(in units) vs. Time

Scatter Diagram 160

Actual Period 140

value (or) number
Y (or) X 120
74 1995
79 1996 100
80 1997
90 1998 80 Period number (or) X
105 1999
142 2000 60
122 2001
40
20
0
1994 1996 1998 2000 2002
Least Squares for Linear Regression
Midwestern Manufacturing
Least Squares Method
Values of Dependent Variables

Time
Least Squares Method
^
Y  a  bX
Where

^
Y = predicted value of the dependent variable (demand)

X = value of the independent

variable (time)

_ _
a = Y-axis intercept
[ XY - n X Y ]
b=  2
_
b = slope of the regression line

2
X - n X 
 
Linear Trend Data & Error Analysis
Midwestern Manufacturing Company
Forecasting Linear trend analysis
Enter the actual values in cells shaded YELLOW. Enter new time period at the bottom to forecast

Input Data Forecast Error Analysis

Actual value Period number Absolute Squared Absolute
Period (or) Y (or) X Forecast Error error error % error
Year 1 74 1 67.250 6.750 6.750 45.563 9.12%
Year 2 79 2 77.786 1.214 1.214 1.474 1.54%
Year 3 80 3 88.321 -8.321 8.321 69.246 10.40%
Year 4 90 4 98.857 -8.857 8.857 78.449 9.84%
Year 5 105 5 109.393 -4.393 4.393 19.297 4.18%
Year 6 142 6 119.929 22.071 22.071 487.148 15.54%
Year 7 122 7 130.464 -8.464 8.464 71.644 6.94%
Average 8.582 110.403 8.22%
Intercept 56.714 MAD MSE MAPE
Slope 10.536

Next period 141.000 8

Least Squares Graph
Trend Analysis

160

140

y = 10.536x + 56.714
120

100
Value

80

60

40

20

0
1 2 3 4 5 6 7
Time

Actual values Linear (Actual values)

 Seasonality Analysis
Ratio = demand / average demand
ler Supplies
Average Seasonal
Demand Demand Ratio Index
80 94 0.851 0.957 83.55556 1 Seasonal Index – ratio of the
75 94 0.798 0.851 88.125 2
average value of the item in a
80 94 0.851 0.904 88.47059 3
90 94 0.957 1.064 84.6 4 season to the overall average
115 94 1.223 1.309 87.88618 5 annual value.
110 94 1.170 1.223 89.91304 6
100 94 1.064 1.117 89.52381 7
90 94 0.957 1.064 84.6 8 Example: average of year 1
85 94 0.904 0.957 88.77778 9 January ratio to year 2 January
75 94 0.798 0.851 88.125 10
75 94 0.798 0.851 88.125 11
ratio.
80 94 0.851 0.851 94 12 (0.851 + 1.064)/2 = 0.957
100 94 1.064 104.4444 13
85 94 0.904 99.875 14
90 94 0.957 99.52941 15
110 94 1.170 103.4 16 If the forecast for January of year 3 is
131 94 1.394 100.1138 17
120 94 1.277 98.08696 18 103, then
110 94 1.170 98.47619 19 Forecast demand Year 3 January:
110 94 1.170 103.4 20 103 X 0.957 = 99 units
95 94 1.011 99.22222 21 Forecast demand Year 3 May:
85 94 0.904 99.875 22
85 94 0.904 99.875 23
106 X130.861
1.309 = 139 units
80 94 0.851 94 24
Deseasonalized Data
 Going back to the conceptual model, solve
for trend:
 Trend = Y / Season
(96 units/ 0.957 = 100.31)
 This eliminates seasonal variation and
isolates the trend
 Now use the Least Squares method to
compute the Trend
Forecast
 Now that we have the Seasonal Indices
and Trend, we can reseasonalize the data
and generate the forecast
 Y = Trend x Seasonal Index
Causal Method
 Y = a + bX as in the least square methods
 Y = the dependent variable
 X = the independent variable
Coefficient of correlation
Causal
student population juice (in glasses)
X Y
1,234 6,042
1,256 5,430
1,345 6,345
1,678 5,671
2,145 6,251
2,345 6,891
2,357 6,789

correlation
coefficient 0.746112
slope = 0.790213
intercept = 4807.424
Multiple regression