When there are outliers in the data set, the robust regression method is a suitable alternative to the conventional regression. Also, if the observations are fuzzy, the conventional regression methods cannot solve the modeling of such observations, and in this case, the fuzzy regression method is a suitable alternative method. For the case where the observations are fuzzy and there are outliers in the data set, the fuzzy robust alternative methods are used. In this paper, for the case where the dependent variables and regression coefficients are fuzzy numbers and the data set contains outliers, the modified fuzzy least squares regression analysis is proposed. In this method, the residuals are ranked to compare the fuzzy sets. The residuals are obtained using the global presence index for each fuzzy set OM. Then, the weight matrix is ​​defined by the membership function of the residuals and the least squares estimates of the weighted fuzzy sets are obtained using the weight matrix. To demonstrate the performance of the proposed method, two examples are presented and their results are presented.