Scribd
Upload
49 views3 pages

Md. Afzal Jamil Solution To The Question No. 4.7 (A, B, C)

The document contains a table with monthly price per chip data, 2 month and 3 month moving averages, and mean absolute deviations (MAD) for each. It then analyzes that the 2 month moving average has a lower MAD, making it a better forecasting method. It also contains exponential smoothing forecasts with different alpha values, finding alpha of 0.5 to have the lowest error.

Uploaded by

afzaljamil
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
49 views3 pages

Md. Afzal Jamil Solution To The Question No. 4.7 (A, B, C)

The document contains a table with monthly price per chip data, 2 month and 3 month moving averages, and mean absolute deviations (MAD) for each. It then analyzes that the 2 month moving average has a lower MAD, making it a better forecasting method. It also contains exponential smoothing forecasts with different alpha values, finding alpha of 0.5 to have the lowest error.

Uploaded by

afzaljamil
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 3

ID:

MD. AFZAL JAMIL 111082

Solution to the question no. 4.7 (a, b, c)

Price per 2 month moving 3 month moving


Month chip($) average (2MMA) average (3MMA) MAD for 2MMA MAD for 3MMA
January 1.8
February 1.67
March 1.7 1.74 0.035
April 1.85 1.69 1.72 0.165 0.12667
May 1.9 1.78 1.74 0.125 0.16000
June 1.87 1.88 1.82 0.005 0.05333
July 1.8 1.89 1.87 0.085 0.07333
August 1.83 1.84 1.86 0.005 0.02667
September 1.7 1.82 1.83 0.115 0.13333
October 1.65 1.77 1.78 0.115 0.12667
November 1.7 1.68 1.73 0.025 0.02667
December 1.75 1.68 1.68 0.075 0.06667
Total = 0.75 Total= 0.793333
        MAD = 0.075 MAD = 0.088

The total number of observations (n)=10 for 2 month moving average and total absolute
error = 0.75
So mean absolute deviation (MAD) for two month moving average = 0.75/10= 0.075
The total number of observations (n)=9 for 3 month moving average and total absolute
error = 0.793333
So mean absolute deviation (MAD) for three month moving average = 0.79333/9 =
0.088

Graph for 4.7 (a, b)

4.7.C
Using Mean Average deviation we find that for 2 months moving average the MAD is
lower than with that of 3 months moving average. That’s why 2 months moving average
is better way of forecasting.
ID:
MD. AFZAL JAMIL 111082

4.7.d

Price per Forecast Forecast with Forecast with Absolute With With
Month chip($) with α=.1 α=.3 α=.5 with α=.1 α=.3 α=.5
January 1.8 1.800 1.800 1.800 0.0000 0.0000 0.0000
February 1.67 1.800 1.800 1.800 0.1300 0.1300 0.1300
March 1.7 1.787 1.761 1.735 0.0870 0.0610 0.0350
April 1.85 1.778 1.743 1.718 0.0717 0.1073 0.1325
May 1.9 1.785 1.775 1.784 0.1145 0.1251 0.1163
June 1.87 1.797 1.812 1.842 0.0731 0.0576 0.0281
July 1.8 1.804 1.830 1.856 0.0042 0.0297 0.0559
August 1.83 1.804 1.821 1.828 0.0262 0.0092 0.00203
September 1.7 1.806 1.824 1.829 0.1064 0.1236 0.1290
October 1.65 1.796 1.786 1.764 0.1458 0.1365 0.1145
November 1.7 1.781 1.746 1.707 0.0812 0.0455 0.0072
December 1.75 1.773 1.732 1.704 0.0231 0.0181 0.0464
Summation
of Errors 0.8632 0.8436 0.7969
Number of
observation 12 MAD 0.0719 0.0703 0.0664

MAD with α = 0.10,


MAD = ∑Absolute Deviation/n
= 0.8632/12 =0.0719

MAD with α = 0.30,


MAD = ∑Absolute Deviation/n
= 0.8436/12 = 0.0703

MAD with α = 0.50,


MAD = ∑Absolute Deviation/n
= 0.7969/12 = 0.0664

On the basis of the analysis, a smoothing α = 0.50 is preferable because it’s MAD is smaller.
Since, there is less error when α = 0.50, the forecasting will be predicted good when there is
less error.

Solution to the problem no. 4.17

Smoothed Forecast Ft=Ft-1+α (At-1-Ft-1)


Smoothed trend Tt= β(Ft-Ft-1)+(1- β)Tt-1 also given α=0.1 and β=0.2

Month Time Actual Smoothed Smoothed Forecast Including


Income (in Forecast Ft (in Trend Tt (in Trend FITt (in
thousands) thousands) thousands) thousands)
Feb 1 70.00 65 0 65
March 2 68.5 65.5 0.1 65.6
April 3 64.8 65.89 0.158 66.05
May 4 71.7 65.92 0.13304 66.05
June 5 71.3 66.62 0.245916 66.87
July 6 72.8 67.309 0.334585 67.64
Aug 7 ? 68.16002 0.437696 68.59
ID:
MD. AFZAL JAMIL 111082

Calculations:
Forecasted Including Trend for August = Exponential Smoothed Forecast (F t)
+Exponential smoothed trend (Tt)

FMarch = 0.1(70)+(1-0.1)(65+0)=65.5
TMarch = β(Ft-Ft-1)+(1-β) Tt-1
= 0.2 (65.5-65)+0.8(0) = 0.1
FApril = 0.1(68.5)+(1-0.1)(65.5+0.1) = 65.89
TApril = 0.2(65.89-65.5)+0.8(0.1) = 0.158
FMay = 0.1(64.8)+0.9(65.89+0.158) = 65.92
TMay = 0.2(65.92-65.89)+0.8(0.158) = 0.13304
FJune = 0.1(71.7)+0.9(65.92+0.13304) = 66.62
TJune = 0.2(66.62-65.92)+0.8(0.13304) = 0.245916
FJuly = 0.1(71.3)+0.9(66.62+0.245916) = 67.30
TJuly = 0.2(67.30-66.62)+0.8(0.245916) = 0.334585
FAugust = 0.1(72.8)+0.9(67.309+0.334585) = 68.16002
TAugust = 0.2(68.16002-67.309)+0.8(0.334585) = 0.437696

You might also like