Abstract:
A major problem for many organisational forecasters is to choose the appropriate forecasting method for a large number of data series. Model selection aims to identify the best method of forecasting for an individual series within the data set. Various selection rules have been proposed in order to enhance forecasting accuracy and good bias. In order to assess the efficacy of selection models in the cases considered, simple selection rules are proposed, based on within-sample best fit or best forecasting performance for different forecast horizons. The current forecasting method that PT.X implement in the past years are using the sales and operation forecast as one of its constrain that lead to an overstock in shampoo X inventory up to 21 percent with amount of 3,351,250 units from its total demand and reduce its liquidity that has an impact to the company cash to producing the next production. Using ARIMA model forecast by analysing the historical data can improve the forecast accuracy from 74% until 85% and also the percentage of overstock from 21% until 0.01%. The current model of the system is not good enough since the frequency of order are often, it needs a new proposed inventory model using economic order quantity that has been reduce the total inventory cost approximately IDR 5,137,452,281.
