EFFICIENCY COMPARISON OF TIME SERIES MODEL CONSTRUCTING BY BOX-JENKINS, SINGLE EXPONENTIAL SMOOTHING AND COMBINED METHOD FOR FORECASTING MONTHLY EGG PRICE

Abstract

The objective of this research was to compare the efficient of time series model were constructing by Box-Jenkins, single exponential smoothing, and combined method for forecasting monthly egg price by the criterion of the lowest mean absolute percentage error. The set of time series data used in research were secondary of 168-monthly egg price collected by the office of agricultural economics during January, 2005 to December, 2018 (Bath per 100 eggs). Data analysis used applied on packaged of MINITAB program version 16.

          The result of this research showed that; the Box-Jenkins model was ARIMA (1, 1, 1) represented by =  when and were forecasting and error value at time respectively, the single exponential smoothing model represented by =  when  and were forecasting and observed value at time respectively, and combined forecasting model represented by =  when ,  and were forecasting value by combined, Box-Jenkins, and single exponential smoothing method at time respectively and MAPE were 5.94%, 2.20% and 5.64% respectively. Hence, the most efficiency model for
forecasting monthly egg price was time series model that construct by single exponential smoothing method.