Estimasi Koefisien Acak Model Vektor Autoregresi

Abstract

The Least Absolute Value (LAV) has long been viewed as an attractive to OLS regression, however, in contract with OLS regression, there are so formulas for estimating the slope and intercept the LAV regression line. Several algorithm exist for these estimate in particular, the linear (goal) programming approach has received much attention. Regression lines estimated by ordinary least square are more severely affected by outliner or extreme data points, than those estimated with the LAV model