Texas A & M University Commerce Plot the Time Series Data Excel Task
Description
4 attachmentsSlide 1 of 4attachment_1attachment_1attachment_2attachment_2attachment_3attachment_3attachment_4attachment_4.slider-slide > img { width: 100%; display: block; }
.slider-slide > img:focus { margin: auto; }
Unformatted Attachment Preview
Single Exponential Smoothing
?t+1 = ?Yt + (1- ?) Y
?t, where the smoothing constant is a positive fraction 0 < ? < 1. Weighted average of
Y
current actual and current forecast is next periods forecast with exponential smoothing factor proving the
weight.
Rearranging we get, ?
Yt+1 = ?
Yt + ? (Yt - ?
Yt). In other words, next periods forecast is current forecast plus
the correction factor proportional to the current error in forecasting. Thus, successive forecasts are based on
the learning/adjustment process based on the errors made in forecasting. The smoothing factor is arbitrarily
selected. If ? is large (near 1), then large weight is given to recent values and very low weight to distant
past. The smoothing of the series is relatively small (small damping factor). If ? is near zero, significant
weight is given to a series of distant past values and very low weight to recent values. The damping factor
is very large and the smoothing of the series is extreme (damping all fluctuations).
Several smoothing factors can be tried, and the best ? value selected by error comparison. We do so for two
values below using Excel Exponential smoothing function: Data? Data analysis ? Exponential smoothing.
Fill the dialogue box for ? = 0.1 with damping factor 1 - ? = 0.9. For ? 0.7 the damping factor is 0.3. The
plots the two forecasts follow (select chart in the dialogue box). The second period forecast is simply taken
as the first year actual by the Excel program. We can use the average of the series of average of first few
observations as the forecast for the first period or use a forecast for this period available from previous
research. But the influence of the choice of initial forecast dies down after a few periods. The sum of
Squares at the bottom is obtained using the Excel Sumsq function.
Year-q
2000-q1
2000-q2
2000-q3
2000-q4
2001-q1
2001-q2
2001-q3
2001-q4
2002-q1
2002-q2
2002-q3
2002-q4
2003-q1
2003-q2
2003-q3
2003-q4
2004-q1
2004-q2
2004-q3
2004-q4
Period (t)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Yt
50
35
25
40
45
35
20
30
35
20
15
40
55
35
25
55
55
40
35
60
Yhat(0.1)
et(.1)
50.00
48.50
46.15
45.54
45.48
44.43
41.99
40.79
40.21
38.19
35.87
36.28
38.16
37.84
36.56
38.40
40.06
40.05
39.55
-15.00
-23.50
-6.15
-0.53
-10.48
-24.43
-11.99
-5.79
-20.21
-23.19
4.13
18.72
-3.16
-12.84
18.44
16.60
-0.06
-5.05
20.45
?
Y (0.7)
50
39.5
29.35
36.81
42.54
37.26
25.18
28.55
33.07
23.92
17.68
33.30
48.49
39.05
29.21
47.26
52.68
43.80
37.64
et(.7)
-15
-14.5
10.65
8.20
-7.54
-17.26
4.82
6.45
-13.07
-8.92
22.32
21.70
-13.49
-14.05
25.79
7.74
-12.68
-8.80
22.36
2005-q1
2005-q2
2005-q3
2005-q4
2006-q1
21
22
23
24
25
75
50
40
65
85
41.59
44.93
45.44
44.90
46.91
33.41
53.29
21.71
5.07
68.49
-18.49
-5.44
55.55
-15.55
20.10
44.66
20.34
38.09
58.90
26.10
7273.85
6246.7
Sumofsq
SumofSq
MSE= 7273.85/24 =
MSE = 6246.7/24
303
260
Exponential Smoothing (0.1)
Value
100
50
Actual
Forecast
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Data Point
error plot (0.1)
50.00
40.00
30.00
20.00
10.00
0.00
-10.00 0
-20.00
-30.00
5
10
15
20
25
30
Exponential Smoothing (0.7)
100
Value
80
60
Actual
40
Forecast
20
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Data Point
et(.7)
30
20
10
0
-10
0
5
10
15
20
25
30
-20
-30
By comparing the errors from the two forecasts we find the smoothing factor 0.7 not only captures the
fluctuations in the data much better but also has lower MSE. Moreover, the errors for ? = 0.1 have a strange
pattern, mostly negative for the first half and mostly positive for the second half. On the other, hand the
errors for ? = 0.7 seem to have relatively random fluctuations around zero, although the variance seems
increasing towards the end. In Excel we can use Solver program to search for the ? value which produces
the smallest overall error (like MSE). I will skip this topic here.
Total points 125. Submit through D2L in Word/Excel format, single or multiple files. Must show
your work, answer all parts of the question, and write necessary explanations to earn full points.
You have to use Excel to solve this numerical problem (project).
Part (A)- 65 points
The data below is for 20 periods of sales in $millions.
t
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Sales(Y)
10
11
14
14
12
15
18
16
16
17
16
20
24
22
20
24
26
28
28
26
Plot the time series data. Forecast for the 21st period using simple exponential smoothing with
alpha 0.3 and 0.8 and plot the actual 20 period values with forecasted values. Compare the two
forecasts using MSE and RMSE. Also calculate the errors and plot the errors for the two
forecasts. Comment on the plots (about stationarity). Do you see systematic underestimation?
Now perform the forecast using Holts two parameter model with alpha = 0.3 and beta = 0.6.
Plot the Holts forecast with actual values. Also calculate the MSE for Holts forecast. Is it better
than the simple exponential smoothing forecasts? Plot the errors for Holts model. Does it look
nearly stationary.
Part (B)- 60 points
For the data below (Production = Y in million tons) plot the data and comment whether you see a
seasonal pattern. Perform Holt-Winters seasonal forecast using alpha = 0.5, beta = 0.4 and
gamma = 0.8. Make forecasts for 2007-q1 and 2007-q2 using Holt-Winters model. Calculate
the errors and MSE and RMSE. Plot the errors and comment on the plot.
Year-q
2001-1
2001-2
2001-3
2001-4
2002-1
2002-2
2002-3
2002-4
2003-1
2003-2
2003-3
2003-4
2004-1
2004-2
2004-3
2004-4
2005-1
2005-2
2005-3
2005-4
2006-1
2006-2
2006-3
2006-4
2007-1
2007-2
period=t
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Y
79
105
95
89
95
137
132
116
115
153
142
127
142
201
179
164
200
243
230
216
222
280
265
255
Holts Double Exponential Smoothing
Holts (Two-parameter) Linear Double Exponential Smoothing Method
Holts (two-parameter) smoothing technique is designed to better handle data with well- defined (significant) trends compared to
the single (simple) exponential smoothing. Holts method allows for evolving local linear trends in a time series. Holts technique
smooths the level and slope directly by using different smoothing constant for each. These smoothing constants provide estimates
of level (Lt) and slope (Tt) that adapt over time as new observations become available. Holts technique provides a great deal of
flexibility (compared to other methods like Browns double exponential smoothing) in selecting the rates of tracking the level and
trend. Three equations are used in this method:
(1) Current level estimate Lt = ?Yt + (1- ?)(Lt-1 + Tt-1); (2) Trend estimate Tt = ?(Lt Lt-1) + (1- ?) Tt-1
?t+p = Lt + p Tt, where Lt = estimate of current level
(3) The forecast for p periods into the future Y
The ? is the smoothing constant for level (0 < ? < 1) and ? is the smoothing constant for trend (0 < ? < 1).
Tt = the trend estimate, Yt = actual observation at period t, and ?
Yt+p = forecast for p periods into the future
Let us use the same example we used for single smoothing with ? = 0.4 and ? = 0.3 (I typed these values in cells J4 and J5).
A
B
C
Year
Quarters (t)
Yt
2000
1
50
2
35
3
25
4
40
5
45
6
35
7
20
8
30
9
35
10
20
11
15
12
40
13
55
14
35
15
25
16
55
17
55
18
40
19
35
20
60
21
75
22
50
23
40
24
65
25
85
2001
2002
2003
2004
2005
2006
D
E
F
G
J
Yt
90
80
70
60
50
40
30
20
10
0
0
5
10
15
20
25
30
Initial values of L and T
To begin the algorithm, the initial L1 is generally set as Y1, and the initial trend value T1 is
[(Y2 Y1)/2 + (Y4 -Y3)/2] = 0 (which happens to be 0 for the given data, but can be non-zero in other data).
The Excel formula I used is =(C3-C2)/2+(C5-C4)/2
The first row has labels and data starts from second row. Therefore, L2 appears in the third row and L1 and
T1 appear in second row.
The Algorithm
Now the algorithm can start according to the equations 1-3 above.
L2 = 0.4(Y2) + (1-0.4)(L2-1 + T2-1) = 0.4(35) + 0.6(50 + 0) = 44?The Excel formula is =$J$4*C3+(1-$J$4)*(D2+E2)
T2 = ?(Lt Lt-1) + (1- ?) Tt-1 = 0.3(L2 L1) + 0.7T1 = 0.3(44 - 50) + 0.7(0) = -1.8? The Excel formula is
=($J$5*(D3-D2)+(1-$J$5)*E2)
?t+p = Lt + p Tt for one period ahead forecast, p = 1 and Y
?t+p = Lt + Tt
Y
?2 = L1 + T1 = 50+ 0 = 50 ? the Excel formula is =D2+E2
Y
?2 = 35 50 = -15 ? the Excel formula is =C3-F3
e2 = Y2 - Y
Successive values are calculated in a similar fashion by copying the formula in successive rows. You can copy all four
formulas together by selecting the four cells in the third row and dragging down the corner. The result is the table
below:
A
B
C
D
E
F
G
Year
Quarters (t)
Yt
Lt
Tt
Yhat(t+p,
p= 1)
et
2000
1
50
50.00
0.00
50.00
0.00
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
35
25
40
45
35
20
30
35
20
15
40
55
35
25
55
55
40
35
60
75
50
40
65
85
44.00
35.32
34.87
37.22
35.56
28.50
27.24
28.82
24.51
19.38
25.78
37.32
38.37
34.59
43.18
49.75
48.32
44.46
51.01
62.02
60.19
53.86
58.40
69.92
-1.80
-3.86
-2.84
-1.28
-1.40
-3.10
-2.54
-1.31
-2.21
-3.09
-0.24
3.30
2.62
0.70
3.07
4.12
2.45
0.56
2.36
4.95
2.92
0.14
1.46
4.48
50.00
42.20
31.46
32.03
35.94
34.17
25.40
24.70
27.51
22.30
16.29
25.54
40.62
40.99
35.30
46.25
53.86
50.77
45.02
53.37
66.98
63.10
54.01
59.87
74.40
-15.00
-17.20
8.54
12.97
-0.94
-14.17
4.60
10.30
-7.51
-7.30
23.71
29.46
-5.62
-15.99
19.70
8.75
-13.86
-15.77
14.98
21.63
-16.98
-23.10
10.99
25.13
6090.21
SumofSq
MSE
243.61
2001
2002
2003
2004
2005
2006
H
I
J
Actual vs Yhat
90
80
70
60
50
40
30
20
10
0
0
5
10
15
Yt
20
25
30
Yhat(t+p)
et
40.00
30.00
20.00
10.00
0.00
0
-10.00
-20.00
-30.00
5
10
15
20
25
30
The Holt-Winters seasonal exponential Smoothing
This procedure allows for both trend and seasonal patterns of the data to be considered and is an extension
of the Holts model. An additional equation is used for the seasonality observed in the data. The model is
helpful in making monthly and quarterly forecasts for inventory management purposes.
Yt
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0
5
10
15
20
25
30
The four equations for multiplicative form Yt=St*Tt*Ct*It of Winters model are:
(1) The level estimate: Lt = ?Yt/St-s + (1- ?)(Lt-1 + Tt-1)
(2) The trend estimate: Tt = ?(Lt Lt-1) + (1- ?) Tt-1
(3) The seasonality estimate: St = ?(Yt/Lt) + (1- ?)St-s
(4) The p periods into future forecast ?
Y t+p = (Lt + pTt)St-s+p
Where the new elements added to Holts model are: ? = the smoothing constant for the seasonality
estimate; St = the seasonal estimate; and s = the length of seasonality
Initial Values
As in the Holts model, the smoothing parameters ?, ? and ? are arbitrarily selected. Let us use values 0.7,
0.3 and 0.4 for ?, ? and ?, respectively, on the above quarterly data. I found these parameters to be the
almost the best using Solver optimization to minimize RMSE). A simple approach is to start the
algorithm after period 5 (for quarterly data). The length of seasonality for quarterly data is 4. The seasonal
indexes for first four quarters can be found by dividing Y1 to Y4 by the average of Y1 to Y4. Then, L5 =
Y5/S1 using =C6/F2 and T5 = (Y5/S1) (Y4/S4)
The Algorithm Starting from S5, L6 and T6
Now the algorithm can start for the next Lt, Tt and St values as per the above formulas. Note that I typed
the three parameter values on cells J1, J2 and J3.
S5 = ?(Y5/L5) + (1- ?)S5-4 = ?(Y5/L5) + (1- ?)S1 = 0.4(Y5/L5) + (1- 0.4)S1 = using Excel formula
=$J$3*(C6/D6)+1-$J$3)*F2 equals 1.33. Next,
L6 = 0.7(Y6/S2) + 0.3(L5 + T5) using Excel formula =$J$1*(C7/F3)+(1-$J$1)*(D6+E6) equals 35.25.
T6 = 0.3(L6 - L5) + 0.7T5 using the Excel formula =$J$2*(D7-D6)+(1-$J$2)*E6 equals -2.18.
S6 = 0.4(Y6/L6) + 0.6S2 using the Excel formula =$J$3*(C7/D7)+(1-$J$3)*F3 equals 0.96.
? t+p = (Lt + pTt)St-s+p gives (for p = 1):
Y
? 5+1 = Y
?6 = (L5 + T5)S2 using the Excel formula =(D6+E6)*F3 equals 28. Finally, e6 = Y6 - Y
?6 using the
Y
Excel formula =C7-G7 equals 7.
Now we can copy the formulas for the next quarters by dragging down (can drag cell formulas together).
1
There are other (more complicated) methods for finding the initial values such as using Regression and
dummy variable Regression, but we will skip those (dealt in later chapters) because, ultimately, the
significance of the choice of initial values dies down as the updating progresses in the algorithm. Besides,
our concern is about the future forecasted values rather than the initial forecasts. After the forecasting we
can calculate the RMSE using =SQRT(SUMSQ(H7:H26)/COUNT(H7:H26)) which gives the RMSE = 6.86.
The forecast for beyond sample periods (second and third quarters of 2006) equal 58.93 and 45.33. The
plots of Holt-winters forecast vs actual values and the plot of errors are given after the Table of
calculations and results.
COLUMNs
A
B
C
D
Lt
E
F
Tt
St
G
?t+p = (Lt+pTt)St+1-p
??
H
I
J
et
?
0.7
Year
Period(t)
Yt
2000-q1
1
50
1.33
?
0.3
q2
2
35
0.93
?
0.4
q3
3
25
0.67
q4
4
40
1.07
2001-q1
5
45
33.75
-3.75
1.33
q2
6
35
35.25
-2.18
0.96
28.00
7.00
q3
7
20
30.92
-2.82
0.66
22.05
-2.05
q4
8
30
28.12
-2.82
1.07
29.98
0.02
2002-q1
9
35
25.97
-2.62
1.34
33.74
1.26
q2
10
20
21.63
-3.13
0.94
22.35
-2.35
q3
11
15
21.49
-2.23
0.67
12.19
2.81
q4
12
40
32.02
1.60
1.14
20.54
19.46
2003-q1
13
55
38.83
3.16
1.37
45.02
9.98
q2
14
35
38.55
2.13
0.93
39.65
-4.65
q3
15
25
38.15
1.37
0.67
27.43
-2.43
q4
16
55
45.64
3.20
1.17
45.04
9.96
2004-q1
17
55
42.75
1.38
1.34
66.91
-11.91
q2
18
40
43.36
1.15
0.93
41.03
-1.03
q3
19
35
50.10
2.82
0.68
29.67
5.33
q4
20
60
51.90
2.52
1.16
61.70
-1.70
2005-q1
21
75
55.60
2.87
1.34
72.73
2.27
q2
22
50
55.31
1.92
0.92
54.20
-4.20
q3
23
40
58.37
2.27
0.68
38.89
1.11
q4
24
65
57.35
1.28
1.15
70.46
-5.46
2006-q1
25
85
61.94
2.27
1.35
78.65
6.35
q2
58.93
Q3
45.33
6.86
RMSE
2
Actual and Winters' Forecast
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0
5
10
15
Yt
20
Yhat(t+p)
25
30
Linear (Yt )
Error Plot of Winters' Forecast
25.00
20.00
15.00
10.00
5.00
0.00
-5.00
0
5
10
15
-10.00
-15.00
3
20
25
Purchase answer to see full
attachment
Explanation & Answer:
Data Excel Task
Tags:
algorithm
data structures
forecasted values
User generated content is uploaded by users for the purposes of learning and should be used following Studypool's honor code & terms of service.