A novel fuzzy regression modeling approach for forcasting purposes in fluctuating conditions

This paper deals with the linear fuzzy regression in which classical regression models in statistics are augmented with the fuzzy sets theory to model the regression relationship. A neglected issue of regression modeling is stochastic fluctuating nature of independent variables which makes the resulted regression models unreliable. To overcome the above issue, a weighted multi-objective optimization model is developed in this paper in order to balance the effects of independent variables in the final regression model. Finally, a case study of global oil price is adopted to validate the developed model in comparison with the existing models.