Beef Price Forecasting based on Temporal, Spatial and Space-Time Parameter Indices
Syifa Nurul Fatimah1 , Ahmad Fuad Zainuddin2,4 , Novi Mardiana2,5 , Utriweni Mukhaiyar3

1)Master Program of Mathematics, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Jalan Ganesha No. 10, Bandung 40132, Indonesia, syifanf34[at]gmail.com
2)Doctoral Program of Mathematics, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Jalan Ganesha No. 10, Bandung 40132, Indonesia
3)Statistics Research Group, Faculty of Mathematics and Natural Sciences, Institut Teknologi Bandung, Jalan Ganesha No. 10, Bandung 40132, Indonesia, utriweni[at]itb.ac.id16
4)Business Mathematics Department, School of STEM, Universitas Prasetiya Mulya, Jalan BSD Raya No. 1, Tangerang 15339, Indonesia, ahmadfuadzain[at]gmail.com
5)Industrial Engineering Department, Engineering Faculty, Universitas Sangga Buana, Jalan PHH Mustofa No. 68, Bandung 40124, Indonesia, novi.mardiana[at]usbypkp.ac.id

Abstract

Beef is among the most sought-after commodities in Indonesia, resulting in significant price fluctuations, particularly during periods such as religious holidays. These price variations affect inflation and necessitate adjustments in government policies concerning beef distribution and imports. Therefore, it is essential to analyze and predict beef prices using empirical data from regions with the highest levels of beef production and consumption. This study aims to examine beef price data through the lenses of temporal, spatial, and space-time dependencies within Java. The methodologies employed in this research include ARIMA, Semivariogram, Kriging, and GSTAR models applied to weekly beef price data from Java.

The findings of this study reveal that beef price fluctuations in Java are primarily influenced by temporal factors, particularly major religious holidays, rather than by location or a combination of location and time. However, there are spatial variations in beef prices across different observation locations. The best predictive model for forecasting beef prices is the ARIMA model. These results provide valuable insights into the patterns of beef prices based on temporal, spatial, and space-time parameters, offering a robust framework for understanding and anticipating price dynamics in the region.

Keywords: beef price, ARIMA, kriging, GSTAR, forecasting

Topic: Probability and Stochastic Analysis