SOLUSI KUADRAT TERKECIL MODEL REGRESI FUZZY DENGAN VARIABEL DEPENDEN FUZZY SIMETRIS

Abstract

ABSTRACT: It is known that the regression model used to analyze the crisp data. In the condition that there is  uncertainty  (imprecision,  vagueness)  on  values  of  observed  variables,  then  the  fuzzy  regression model is required. In this paper discussed a fuzzy regression model with symmetrical fuzzy dependent variable and crisp independent variables using the least squares approach. The method used to find the  linear  model  is  to  minimize  the  distance  function  between  observed  fuzzy  variables  with  the estimation dependent variable or the corresponding theoretical value. It is shown that the solution of this  model  is  a  generalization  of  the  classical  regression  model. Further  discussion  is  about  the properties of solution of the model.

Keyword: center model, spread model, fuzzy distance, iterative solution.