COMPARE BETWEEN ORDINARY LEAST SQUARE AND MAXIMUM LIKELIHOOD METHODS FOR ESTIMATE PARAMETER OF FUZZY SPATIAL LAG MODEL

Abstract

This paper deals with the study about the formulation of Spatial Lag model for independent and dependent fuzzy variables. while the parameters crisp values, with compare between ordinary least square (OLS) and maximum likelihood (MLE) methods to Estimate Parameters by Criteria Root Mean Squares Error ( RMSE) and Mean Absolute Percentage Error (MAPE) ,we get results that (OLS) the best from (MLS) for Trapezoidal fuzzy number in the domain traffic accidents for a number of cities in Iraq for the year 2018.and that after converting the Trapezoidal fuzzy number into crisp values by centriod method , calculations the results by Matlab language.