Optimization ELM neural network in prediction problem: case study forecasting demand steel in Iran

Subject Areas : مدیریت

Jalal  Rezaeenour ¹ (Associate Professor, Department of Industrial Engineering, Qom University, Qom, Iran)
Mansoureh  Yari eili ² (Ph.D. Student, Department of Computer Engineering and Information Technology, Qom University, Qom, Iran)
Esmaiel  roozbahani ³ (Assistant Professor, Department of Industrial and Computer Engineering, Birjand University of Technology, Birjand, Iran)
Mohammad hossein  Roozbahani ⁴ (PhD student ,Department of Mechanical Engineering, Tarbiat Modarres University, Tehran, Iran)

Received: 2019-09-18 Accepted : 2019-09-18 Published : 2019-08-23

Keywords: Genetic Algorithm, Artificial Neural Network, Extreme learning machine, Supply &amp, demand steel Prediction,

Abstract :

Prediction of supply and demand, is a crucial issue for supply chain partners. With the accurate forecasted supply and patterns that indicate the sizes and directions of future production flow, the government and suppliers can have a well-organized strategy and provide a better infrastructure for improving industrial sector.With the aim of developing accurate forecasting tool in steel industry, this study present a new optimized neural network, by combination of Extreme Learning Machine and genetic algorithm. We employed our model on a dataset for steel supply - demand in Iran from jul-2009 to jan2013 to estimating the performance. The results show that prediction accuracy and performance relatively better than other classical approaches, according to RMSE and MAPE evaluations. In our model. Based on statistical tests, our new model is better than other model in accuracy, so can be used in as a suitable forecasting tool in steel supply forecasting problems.