Using linear programming method to estimate parameters of linear regression model by absolute deviation

This work deals with descriptive summary of linear programming subject, and describes least square method for estimating regression parameters. As well as using other methods in order to estimate the parameters unconventional on reduced variance (or reduce sum of square error) , but depend on minimizing variation of absolute values from the median , the aim of this work is to find an easy and precise way to estimate the absolute deviation , which is important when the variance fail in precise estimate and lead to enlarge the variation of the data , while least square method depend on minimizing sum of the square , therefore , the goal is to reduce the deviation of the absolute values from the mean in model of linear programming in order to estimate absolute deviation which is a simple and precise method , and this manuscript includes examples of this application.