In real world life, the major difficulty faced by the decision maker is to forecast the demand,
since there usually involves uncertainty as to how many items customers will buy during the single
period. At the end of the period, the items become obsolete. The existing literature discusses
the models with probabilistic demand [4-61, in that probability distributions are estimated from
historical data. However, in many cases, especially for new product,, the probabilities are not
known due to lack of historical data. In this case, the demands are more suitably described by
linguistic terms such as high and low, or are approximately estimated to be equal to a certain
amount subjectively determined by the expert. Restated, the demands are fuzzy.