Studi model statistika data deret waktu untuk peramalan produksi pangan

Abstract

For casting of food production for one year ahead or more is very important information for maintaining food security of community. By knowing food availabitiy in the future, government can prepare better food and nutrition planning. Therefore the objectives of this study are: 1) to find time series models of some main food production according to location so that the models can be used to forecast food poduction in the future, and 2) to make comparasion the food production models wheter type of food depending on location or not. Autoregressive, Moving Average, Mixed Autoregressive Moving Average and Autoregressive Integrated Moving Average were applied in study to find best model of time series data of food production. This study resulted models that can be used statistically to forecast food production in the future. The model for production of rice, cassava, sweet potato, corn, peanut, soybean, cow, chicken, duck, bufallow, and goat in Subang and Bandung district are ARIMA(p,d,q) where p<l, d<l and q<l. By analysis, it is known that model of spesific food will depend on location.