The point forecasting is a poor decision making aid. The performance of point forecasting reduces when there is an extreme and uncertain situation (Li et al., 2012). Chatfield (2000) mentioned the importance of providing the interval forecasts such as
assessing future uncertainty, planning different strategies for the range of possible outcomes,
comparing forecasts and exploring different scenarios based on different assumptions. In food
retail industry, the estimation of forecast interval is very important for inventory planning to
assess the appropriate levels of safety stock. To overcome this issue, the point forecasts were
converted into the interval forecasts in most of the earlier studies. Granger et al. (1989)
suggested the confidence intervals for 10% and 90% as
ðY t 7 1:282σt Þ. The confidence interval of the prediction is a range
that is likely to contain the mean response for the given values of independent variables in a
model. However, the prediction interval of forecast is a range that is likely to contain the mean
forecasted response for the given value of independent variables in a model (Walpole et al.,
2012). Chatfield (2000) and Hyndman and Atha- nasopoulos (2013) mentioned the calculation of
prediction inter- vals using theoretical formulae that are basically of the same
general form. Prediction interval at 100ð1 - αÞ% for the value h
steps ahead can be represented by