Cracking the Code of Crop Growth: Illuminating the Future of Philippines’ Onion Production for a Resilient Filipino Diet with the ARMA Forecasting Model (Published)
This study employed the Box-Jenkins methodology and the Autoregressive Moving Average (ARMA) model to forecast onion production in the Philippines. By utilizing historical data from the Philippine Statistics Authority, an optimal forecasting solution was achieved through the selection of the ARMA (4,2) model. The model demonstrated a favorable fit, passing diagnostic tests and exhibiting a mean absolute percentage error (MAPE) of 10.406%. Projections for onion production in 2023 and 2024 were provided, highlighting expected yields for each quarter. The analysis of historical data revealed periodic fluctuations in onion supply driven by factors such as weather patterns, market demand, agricultural practices, and imports or exports. The study’s implications emphasize the value of accurate forecasting models for decision-making in production planning, resource allocation, pricing, and market positioning. Policymakers, farmers, and stakeholders can utilize the findings to optimize onion production sustainably and enhance the agricultural sector’s performance in the Philippines.
Keywords: Agriculture, Box-Jenkins analysis, Forecast, Python, autoregressive moving average (ARMA) model, log transformation, onion production, onion supply, variance, variance stabilization