The Box Jenkins ARIMA methodology has been utilized in this study for forecasting winter vegetables in Bangladesh. Time series data are used for this empirical study of modelling some selected winter vegetables The best fitted Box-Jenkins ARIMA model for forecasting Bean, Cabbage and Cauliflower are (0,2,1) (1,2,3) and (0,2,1) respectively. The diagnostic checking of these model is confirmed by both empirical and graphical methods, AIC and BIC of models are observed and the residuals are examined thorough ACF, PACF and QQ plots. Further, the future production of winter vegetables is forecasted up to 10 years, as precise as probable, from the best selected model. The performances of ARIMA models provide reasonable fit to the observed data since the prediction is good during and beyond the assessment period 1961 to 2017. This information of future production would be helpful for policy maker in decision making on necessities of storage facility, import and/or export and so on.
Key words: Bean, Cabbage, Cauliflower, Ljung-Box Test, ADF test
|
