Development of Artificial Neural Network Model for Estimation of Salt Fields Productivity

Indra Cahyadi (Department of Industrial Engineering, Engineering Faculty, Universitas Trunojoyo Madura, Indonesia)
Heri Awalul Ilhamsah (Department of Industrial Engineering, Engineering Faculty, Universitas Trunojoyo Madura, Indonesia)
Ika Deefi Anna (Department of Industrial Engineering, Engineering Faculty, Universitas Trunojoyo Madura, Indonesia)

Jurnal Teknik Industri
Vol. 20 No. 2 (2019): August

In recent years, Indonesia needs import millions of tons of salt to satisfy domestic industries' demand. The production of salt in Indonesia is highly dependent on the weather. Therefore, this article aims to develop a prediction model by examining rainfall, humidity, and wind speed data to estimate salt production. In this research, Artificial Neural Network (ANN) method was used to develop a model based on data collected from Sumenep Madura Indonesia.  The model analysis used the complete experimental factorial design to determine the effect of the ANN parameter differences. Furthermore, the selected model performance compared with the estimate predictor of Holt-Winters. The results presented that ANN-based models were more accurate and efficient for predicting salt field productivity.