Generalized Space Time Autoregressive Forecasting of Chili Prices
Indonesian people are mostly like spicy. Therefore, Chili becomes an ingredient in cooking that cannot be separated from Indonesian people. In certain months there is too much demand from the public and inversely proportional to the stock of chili. So, appreciate the chili prices to rise. In the heart of statistics, several methods can be used as forecasting as well as data series from univariate, bivariate and multivariate cases. The reason for choosing the method is following the analysis needs and available data. One of the popular methods in the multivariate case is Generalized Spacetime Autoregressive (GSTAR) the advantage of this method is that it can capture the characteristics of location-based data as well as time. For this reason, in this paper, we used chili prices data in 4 major cities in Java Province, Indonesia. Such as Jakarta, Bandung, Semarang, and Yogyakarta. As a result, GSTAR (21) I (1with inverse weighted be the best models because it fulfilled white noise and normal multivariate assumption with the lowest RMSE and MAP.
International Conference on Information Management and Technology 2018
Jamilatuzzahro, Rezzy Eko Caraka, Riki Herliansyah, Asmawati S., Dian Megah Sari, Bens Pardamean