Time Series Based Crude Palm Oil Price Forecasting Model with Weather Elements using LSTM Network
Kasturi Kanchymalay1, N. Salim2, Ramesh Krishnan3

1Kasturi Kanchymalay*, Faculty of Information Technology Engineering, University Technology Malaysia, Melaka, Malaysia.
2N. Salim, Faculty of Engineering, School of Computing, University Technology Malaysia, Melaka, Malaysia.
3Ramesh Krishnan, Faculty of Business & Management, University Technology MARA, Malaysia.
Manuscript received on September 21, 2019. | Revised Manuscript received on October 05, 2019. | Manuscript published on October 30, 2019. | PP: 3188-3192 | Volume-9 Issue-1, October 2019 | Retrieval Number: A9994109119/2019©BEIESP | DOI: 10.35940/ijeat.A9994.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: In field of agro economic, Crude Palm Oil (CPO) price forecasting is still heavily relies on human expertise. This paper proposes a CPO price forecasting model to assist the palm oil plantation organization in anticipating more effectively monthly fluctuations and manage the supply and demand efficiently avoid problems of price going very low. The parameters used by the predictor consist of weather variables, namely, temperature, rain amount, pressure, humidity and radiation as well as past CPO price. CPO price for past 10 years collected from MPOC and the environmental parameters collected from meteorology department of Malaysia during the period 2005 to 2016, were used to model CPO price using a Long-Term Short Memory Network (LSTM). Our results showed that the LSTM model predicted monthly fluctuations of the price with an average accuracy of 90%. The contribution suggests that the LSTM based forecasting could assist worldwide palm planters in decision making on palm oil crop management and operation processes.
Keywords: Artificial Neural Network, Forecasting; Machine Learning, Time series, Weather Elements