Professional Documents
Culture Documents
Forecasting
1. Introduction
c Rakesh Nagi
University at Bualo - IE
26
Forecasting
Approaches to Forecasting
Subjective Methods - human judgment
(1)
University at Bualo - IE
27
Forecasting
1. Trend
2. Seasonality
University at Bualo - IE
28
Forecasting
Notation
Di : Demand during period i
Basic idea of TS
Ft;t+ =
c Rakesh Nagi
1
X
n=0
anDt
University at Bualo - IE
(2)
29
2. EVALUATING FORECASTS
2. Evaluating Forecasts
et = error in period t
= Ft ;t Dt
= Ft Dt(onestep)
MAD =
n
X
n i=1
n
X
j
eij
n i=1
(3)
(4)
(5)
University at Bualo - IE
31
Dt = + t
c Rakesh Nagi
University at Bualo - IE
(6)
33
n i=t
(7)
c Rakesh Nagi
University at Bualo - IE
34
Ft = Dt 1 + (1 )Ft
= Ft 1 et 1
Ft
= Dt
Ft = Dt
1
+ (1
1 + (1
2
)Ft 2
)Dt 2 + (1 )2Ft
University at Bualo - IE
35
c Rakesh Nagi
University at Bualo - IE
36
University at Bualo - IE
(8)
38
Y^ = a + bX
(9)
where
n
X
Sxy = n iDi
i=1
2
n2(n + 1)(2n + 1)
Sxx =
D =
c Rakesh Nagi
n
X
n i=1
Di
University at Bualo - IE
(10)
(11)
(12)
n
X
Di
i=1
n2(n + 1)2
4
39
St = Dt + (1 )(St 1 + Gt 1) \intercept"
Gt = (St St 1) + (1 )Gt 1 \slope"
new slope
old slope
Ft = St + Gt (one step)
Ft;t+ = St + Gt ( step)
Usually,
c Rakesh Nagi
University at Bualo - IE
40
c Rakesh Nagi
University at Bualo - IE
42
c Rakesh Nagi
University at Bualo - IE
43