In recent decades regarding to its ability in describing, eva‎luating, modeling and forecasting the response variable within explanatory variables in case of imprecise, noisy, linguistic, qualitative data or vague situations, Fuzzy Linear Regression (FLR) is one of the most popular and widely used fuzzy methods. Regularly, the FLR is one of the best methods of modeling when the number of observations for modeling is low, distributional assumptions are violated, the relationship between input and output variables is vague or the phenomenon is ambiguous. The operation logic in this method is to minimize the vagueness of the model, defined as the sum of individual spreads of the fuzzy parameters. Although this process is coherent and can obtain the narrowest α-cut interval and exceptionally the most accurate results in the training data sets, it cannot guarantee to achieve the desired level of generalization. While the quality of made managerial decisions in the modeling-based field is dependent on the generalization ability of the used mode. On the other hand, the generalization capability of a model is generally dependent on the accuracy as well as reliability of results, simultaneously. A new methodology is proposed for the fuzzy linear regression modeling; in which in contrast to conventional models, the constructed models' reliability is maximized instead of minimizing the vagueness. To comprehensively eva‎luate the proposed model's performance, 74 benchmark data sets are considered from the UCI. Empirical results indicate that, from a general perspective, in 64.86% of cases, the proposed model has better generalization ability. narrower α-cut interval as well as more accurate results in the interval and point estimation, than traditional ones. It is clearly illustrated the importance of the reliability of results in the addition to the accuracy that is considered in none of the conventional FLR modeling procedures. Therefore, the proposed EFLR model can be considered as an appropriate alternative in fuzzy modeling fields, especially when more generalization is desired

مهدي خاشعي آشياني

رضا حجازي طاقانكي , حسين خسروشاهي