Forecasting of Sporadic Demand Patterns with Seasonality and Trend Components: An Empirical Comparison between Holt-Winters and (S)ARIMA Methods

Table 1

Data collection and preliminary analysis.

Series	Group	${\text{CV}}_{31}$	${\text{ADI}}_{31}$	${\text{CV}}_{24}$	${\text{ADI}}_{24}$	ddp	Seasonality	Trend
s1	1	1.80	1.35	1.79	1.26	reject	x
s2		1.54	1.11	1.45	1.09	Geometric ($6.34e-002$)	x
s3		1.30	1.19	1.35	1.09	Neg. Binomial (1, $4.75e-002$)	x
s4		1.09	1.48	1.01	1.33	Neg. Binomial (3, $7.76e-002$)	x
s5		1.22	1.19	1.21	1.14	Neg. Binomial (2, $4.11e-002$)	x
s6		1.25	1	1.30	1	reject	x
s7		2.40	1.29	2.36	1.33	reject	x
s8	2	1.11	1.41	1.33	1.50	reject	x	x
s9		2.38	1.35	1.77	1.14	Neg. Binomial (2, $4.88e-002$)	x	x
s10		1.63	1.41	1.69	1.41	Neg. Binomial (1, $9.88e-002$)	x	x
s11		1.28	1	1.42	1	Neg. Binomial (1, $4.7e-002$)	x	x
s12		1.30	1	1.34	1	Neg. Binomial (2, $1.6e-002$)	x	x