An Adaptive Neural Network-Fuzzy Regression Approach for Optimization of Steel Price Forecasting in Complex and Non-Linear Environment

This paper presents an adaptive neural network (ANN)-fuzzy regression approach for optimization of steel price estimation and forecasting in uncertain, complex and non-linear environments. Steel price is viewed as the resultant of standard economic indicators which are exchange rate, import, export, gross domestic production, oil price, and overall price level. The proposed model is executed in Iran by considering 48 months from 2003 to 2006. Several ANN models are trained and tested and the ANN model with a minimum mean square error (MSE) is compared with seven distinct fuzzy regression approach and conventional regression models based on Mean Absolute Percentage Error (MAPE).

An Adaptive Neural Network-Fuzzy Regression Approach for Optimization of Steel Price Forecasting in Complex and Non-Linear Environment Keywords:

Forecasting , Sensitivity Analysis , Optimization , Artificial Neural Network , Fuzzy Regression