Torsion relative Percentage Bend to estimate model parameters seasonal Aldharba existence values ​​stray

Abstract

Abstract Data collection often contains of outliers that influence not only in model diagnostic which depends on autocorrelation and partial autocorrelation coefficients, but exceed to model parameters estimation, so the goal of this paper is to make the autocorrelation coefficient robust against outliers, thus use the moments estimation method (or so-called Yule-Walker equations) to estimate model parameters which will be robust depending on those coefficients. This goal is achieved by extending the robust Percentage Bend Correlation method to use it in estimate the autocorrelation coefficients which will be used in parameter estimation.