This study was done to determine the trend of banana of two classes namely, Class A and Class B, using the Box and Jenk ins methodology. The monthly production of bananas in boxes was used to an alyze the series. The time series data covers the production from January 2005 to D ecember 2010, where the production in the year 2010 was compared to the pre dicted values for accuracy. The method used ARIMA to model and forecast the banana production for each of the two classes. The identified models for class A and c lass B bananas were MA(12) and ARIMA(1,6,2) respectively, and was identified t o be well fitted to the series showed by some statistical tests. Anderson-Dar ling test was used to check the normality of the residuals and Ljung-box test for t he independence. However, results showed that the models were unable to give accurate forecasts since both models yield large values for MSE and MAPE, which were use d to measure the accuracy of the forecasts. Class A series resulted to MS E with 17,502,323while MAPE resulted to 29.03%, class B, on the other h and, resulted to 235,359.8 for MSE and 26.83% for MAPE. It was concluded that t he number of data points was not able to represent well the production series which led to the large errors in the forecasts.
|Source Institution||National Statistical Coordination Board|
|Place of Publication||Davao City|
|Original Publication Date|