Regression analysis usage in researching tendencies and economical security regularities and criminality

Abstract

This work analysis the linear regression model and conditions which influence the correctness of constructing this model. This work also investigates methods used to estimate the unknown variables of linear regression. The primary purpose of the work is to compare unknown variables estimation methods and their prediction precision for the given problem. The dataset of Boston buildings information was used during the investigations. Given example exceptionally performs this task because it provides an opportunity to estimate the price dependency on various factors on which it depends. In this work, the following unknown variables estimation methods of linear regression are reviewed: least-squares method, most minor absolute deviations method, maximum likelihood estimation and orthogonal regression—algorithms for calculating its indicators are given. The advantages and disadvantages of each technique are explained. Errors of the above-given methods are compared.