Implementation of Backpropagation Artificial Neural Network on Forecasting Export of Palm Oil in Indonesia

Adinda Dwi Putri (Universitas Negeri Padang)
Dina Fitria
Nonong Amalita
Zilrahmi

UNP Journal of Statistics and Data Science
Vol. 1 No. 5 (2023): UNP Journal of Statistics and Data Science

Export activities are one of the largest revenues in Indonesia with the largest contributor to export is being palm oil. Increasing volume of palm oil exports, it will be able to spur economic growth in Indonesia. In this research, palm oil export forecasting in Indonesia is carried out based on the main destination countries using the Artificial Neural Network (ANN) method with the Backpropagation algorithm. The data used is palm oil export data for 2012-2022 obtained from the Central Statistics Agency (BPS) website. From the data used, the optimal architecture model is 10-1-3-3-1 with a MAPE of 9.68%, which means that this architecture uses 10 input data, 3 hidden layers with the number of each input neuron (1,3,3), and there is 1 output output. From this study, it is estimated that 90% of the results of palm oil export forecasting using the ANN method are close to the actual value.