Brent Crude Oil Price Forecasting using the Cascade Forward Neural Network

Brent Crude Oil Price Forecasting using the Cascade Forward Neural Network

artificial neural networks; cascade forwards; brent; crude oil; MAPE

Crude oil is one of the most traded non-food products or commodities in the world. In Indonesia, crude oil will still be a
contributor to the Gross Domestic Product in 2021. Excessive consumption of fuel oil (BBM) in Indonesia has resulted in a
scarcity of crude oil, especially diesel. Forecasting the price of Brent crude oil is an important effort to anticipate fluctuations
in the price of fuel oil. The Cascade Forward Neural Network (CFNN) method is proposed to forecast fuel prices because of
its superiority in fluctuating data types. The data used in this research is the price of Brent crude oil in the period January
2008 to December 2022. The CFNN method will be evaluated using the Mean Absolute Percentage Error (MAPE) to choose
the best architectural model. The best Architectural Model is used to predict the next 12 months. After 10 architectural model
trials, 2-6-1 became the best model with a MAPE data training value of 6.3473% and MAPE data testing of 9.4689%.
Forecasting results for Brent crude oil for the next 12 months tend to experience a downward trend until December 2023.

Fatkhurokhman Fauzi, Dewi Ratnasari Wijaya, Tiani Wahyu Utami

http://jurnal.iaii.or.id

Professional Organization Ikatan Ahli Informatika Indonesia (IAII)/Indonesian Informatics Experts Association

August 2023

Sri Wahyuni

ISSN Media Electronic: 2580-0760

PDF

English

Text