This paper presents the comparison between transfer function method and neutral networks for predicting rubber smoked sheet prices. The two procedures are based on time series data extracted from 2001 to 2003. The accuracy of two forecasting models is evaluated using (1) root mean squared error: RMSE and (2) mean absolute percentage error: MAPE. The results indicate that the traditional statistical approach has a potential to forecast rubber smoked sheet prices compared to neural networks model. In addition, the future directions are also discussed in the paper.
Keywords: Time series, Transfer function method, Neutral networks, Box and Jenkins method, Rubber smoked sheet (RSS3) prices, Back propagation
Corresponding author: E-mail: walailak_ut@yahoo.com , me_porntip@yahoo.com
Chatchaipun, P. ., & Atthirawong*, W. . (2018). Modelling and Forecasting Daily Rubber Smoked Sheet Prices. CURRENT APPLIED SCIENCE AND TECHNOLOGY, 315-320.
