A linked list data structure for a simple linear regression algorithm (Q1085037)

The simple linear regression problem arises from a fundamental model of statistical and economic forecastings. The model consists of an independent (also known as predictor) random variable which is used to determine the value of the dependent (or response) random variable. The problem is to find the linear equation which will fit the data comprising these two variables. The least absolute values estimation is the criterion used here to estimate the values of these two variables by minimizing the sum of the absolute deviations of a given set of observations from the values predicted by the model. This paper presents a special purpose algorithm utilizing linear programming and a special data structure to solve this problem efficiently.