“…Nevertheless, Bootstrap method requires large sample size and a large number of previous statistical material to guarantee the accuracy of the forecast. The theoretical evidence (Porteus, 1985;Silver, Pyke, & Peterson, 1998), the digital simulation (Babai, Jemai, & Dallery, 2011;Sani & Kingsman, 1997) and the case from Porras and Dekker (2008) conformably indicate that the (s, S) system is more proper. However, due to that intermittent demand has little nonzero value, which means it is hard to get accurate distribution of demand, therefore, we can't calculate out s and S; even the distribution is known, the calculation process of optimal (s, S) is complicated.…”