Modelling and Forecasting Natural Rubber Price in Malaysia: Comparison Between HES, ARIMA and ANN

• Published in: Springer Proceedings in Mathematics and Statistics
• Main Author: Shaharuddin H.I.B.; Azmi N.N.B.K.; Shamsudin F.I.M.B.
• Format: Conference paper
• Language: English
• Published: Springer 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85209581857&doi=10.1007%2f978-981-97-3450-4_10&partnerID=40&md5=ca2933214af174f833cb859376e48a5c

Standard Malaysian Rubber 20 is one of the best qualities of natural rubber that contributes to the income of Malaysia since Malaysia is one of the exporters of natural rubber. Modelling and forecasting the price of Standard Malaysian Rubber 20 is the goal of this study. The data used is daily prices per kilogram (kg) in unit cents (Malaysian Ringgit (RM)) of Standard Malaysian Rubber 20 which covered a 300-day period commencing on January 20, 2021, and ending on April 14, 2022. Three methods of forecasting were employed which included Box-Jenkins Model (ARIMA), Holt’s Exponential Smoothing (HES), and Artificial Neural Network (ANN). Additionally, the study seeks to evaluate how well the models perform in relation to the price of Standard Malaysian Rubber 20. The best model obtained was used to forecast the next five periods ahead of the origin. The findings of this study showed that the best models for each method are Holt’s Exponential with the value of α = 0.9999 and β = 0.0006, ARIMA(0,1,1) and ANN with one hidden layer with 3 units. This research found that ANN is the best-fitted model for the price of Standard Malaysian Rubber 20. The forecast values show that the price of Standard Malaysian Rubber 20 is expected to increase in the next 5 consecutive days. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.