Price Prediction of Staples with Simultaneous Fourier Series Estimator as a Contribution to Solve the National Food Problem
DOI:
https://doi.org/10.36481/diujbe.v16i1.gck3n054Keywords:
Fourier series, Staples, Cooking oi, Chili, Chicken eggs, Price prediction, Time seriesAbstract
This study provides an overview related to the prediction of prices for staples that are widely consumed by the Indonesian people and often experience price fluctuations at certain times, namely cooking oil, chili, and chicken eggs. Based on the Market Monitoring System and Staples of the Ministry of Trade, the national average price of cooking oil as of January 3, 2022, reached USD 1.17
per liter. The average price of chicken eggs reached USD 1,98 per kilogram, an increase of 20.16% at the same time. In addition, based on the National Strategic Food Price Information Center, the average price of chili in traditional markets throughout the province rose by USD 0.19 or about 3.16% and reached USD 6.29 per kilogram. The prediction of staple food prices in this study was formed using a nonparametric regression model of the Fourier series estimator. This regression model has high flexibility in forming the model and estimating the fluctuation data pattern, which is not known in the form of time-series data. The data in this study is sourced from the National Strategic Food Price Information Center website with the response variables used being the daily prices of cooking oil, chicken eggs, and chili from September 1, 2021, to December 31, 2021, as training data. In addition, the same commodity data from January 1, 2022, to January 10, 2022, is used as testing data. The best model from the estimation results with the Fourier series is a model with an oscillation parameter of 25 which contains a sine cosine basis with a coefficient of determination of 83.05%
and an MSE value of 1.628. The selected model also has a very good performance predicting the prices of the three staples with a MAPE value of 1.082%.