Piecewise Fuzzy Linear Model with Measurement Error Variable

측정오차가 있는 경우의 분할 퍼지회귀모형

  • 안정용 (서남대학교 전산통계학과) ;
  • 한범수 (서남대학교 전산통계학과) ;
  • 최승현 (서남대학교 전사통계학과)
  • Published : 1995.10.01

Abstract

In this study we present the inverse correlation method to select the exploratory variables, while Sugeno used RC method in his paper[6] We assume linear model with measurement error variables as in Fuller's Book[9]. we provide possibilistic linear model and predict the fuzzy response variable in case of fuzzy exploratory variables. By plotting data we can divide them for piecewise plane and provide the piecwise possibilistic linear model. If the exploratory variable is fuzzy trapezoidal variable or interval variable, then we estimate fuzzy trapezoidal variable or interval variable, then we estimate fuzzy trapezoidal response variable respondent to it. We will illustrate using Nonlinear System data in Sugeno's paper

Keywords