Tesdoo 0
Tesdoo 0
Question 1
(i) Briefly outline the steps required in deseasonalizing a time series using MA
and additive model.
(ii) Data on then following table shows the monthly shipment of air conditioners
of a leading manufacturing company from 1983 to 1986. Complete the table.
(NB the seasonal indices have been scaled by a constant)
1
Aug 121.1 115.83 105 102.5 102.59 118.0
Sep 114.2 116.58 98 98.25 98.33 116.1
Oct 116.1 117.32 99 99.75 99.83 116.3
Nov 118.6 118.06 100 100.75 100.83 117.6
Dec 139.5 118.81 117 120 120.10 116.2
1986 Jan 105.6 119.55 88 89.25 89.32 118.2
Feb 99.7 120.29 83 86 86.07 115.8
Mar 114.2 121.03 94 98 98.08 116.4
Apr 115.7 121.78 95 98.5 98.58 117.4
May 125.4 122.52 102 104 104.09 120.5
Jun 120.4 123.26 98 102.25 102.34 117.7
Jul 120.7 124.01 97 99.75 99.83 120.9
Aug 124.1 124.75 99 102.5 102.59 121.0
Sep 124.6 125.49 99 98.25 98.33 126.7
Oct 123.1 126.24 98 99.75 99.83 123.3
Nov 120.8 126.98 95 100.75 100.83 119.8
Dec 151.5 127.72 119 120 120.10 126.1
(5+10=15 marks)
Question 2
(i) Derive the weights and write down the expression for a simple moving
average of order 2, overlaid by a moving average of order 4.
(ii) State and give a brief description of the four components of a time series data.
(6+9=15 marks)
Question 3
Define a turning point and show that the mean of the total number of turning points is
2
(n - 2) . (10 marks)
3