Académique Documents
Professionnel Documents
Culture Documents
Series
Analysis:
Time Series
model
Analysis
includedecomposition
two steps:
Identify factors which influence the variations in
the series.
Isolating, analyzing & measuring the effect of
these factors independently.
Purpose of decomposition is to break a series
into components:
trend value (T), seasonal variations (S), cyclical
Multiplicative
Model
Y= T*C*S*I
This model is effective where the effect of C,S
Additive Model
Y= T+C+S+I
T, C, S, I are absolute quantities and can have
1.FreehandCurveMethod
2.SmoothingMethods:
SemiAverageMethod
MovingAverageMethod
4.TrendProjectionMethods
LeastSquareMethod
1985 86
Prod (lacs) 15
23
87
88
89
90
19
27
24
30
P
R
O
D
U
C
T
I
O
N
YEARS
SEMI-AVERAGE METHOD:
Acctothismethodtheoriginaldataisdividedintotwo
parts.Thenthemeanofboththepartsiscalculated
separately.
Ex2.Determinethetrendbyapplyingthesemi-average
method.
Yr. 1990 91
92
93
94
95
Sales 50
58
54
71
66
74
Table Format
Years Production/Sales.
1
2
3
4
5
6
7
8
X1
X2
X3
X4
X5
X6
X7
X8
3yearly
movingtotal
X1+X2+X3
X2+X3+X4
X3+X4+X5
.
.
.
.
3yearlymoving
average
X1+X2+X3/3
X2+X3+X4/3
X3+X4+X5/3
.
.
.
.
Ex 4Findtheseculartrendby3yearlymoving
average.
Yr
Sales(inlacs)
2001
45
2002
40
2003
48
2004
62
2005
54
2006
75
2007
50
2008
82
1
2
A+B/2
B=X2+X3+X4+X5/4
X4
B+C/2
X3+X4+X5+X6
5
6
A=X1+X2+X3+X4/4
X3
X2+X3+X4+X5
4YearlyCentered
MovingAverage
(Trend)
X5
X6
C=X3+X4+X5+X6/4
Ex5: Findtheseculartrendby4yearlymovingaverage
Yr
Value
2001
506
2002
620
2003
1036
2004
673
2005
588
2006
696
2007
1116
2008
738
2009
663
Trend
Methods
It is bestProjection
represented by straight line is
termed as long run directions(upward,
downward, constant), of any business
activity over a period of several years.
Reasons to study trend:
Helps in describing long term general
XY =
Na+bX
aX+bX2
Function Equation
Y=a+bx
Year
Prod
/Sales
X2
XY
Trend
Values
Yc=
a+bX
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
X2
XY
Yc
Ex 6Fitthestraightlinetrendbyleastsquaremethod:
Yr
Production(lacs)
2002
2003
2004
2005
10
2006
12
2007
17
Ex7 Fitthestraight-linetrendbyleastsquaremethod.
Alsopredictthesalesforyear2010
Yr
Sales(000)
2003
80
2004
90
2005
92
2006
83
2007
94
2008
99
2009
92
Fitthestraight-linetrendbyleastsquaremethodby
assuming2005asthebaseyear.Alsopredictthesales
foryear2010
Yr
Value
2001
506
2002
620
2003
1036
2004
673
2005
588
2006
696
2007
1116
2008
738
2009
663