Research of cigarette sales quantity forecast based on multiple linear regression model

ABSTRACT: In order to determine the main affecting factors for tobacco consumption and the future development trends of the cigarette sales in Guiyang from a macroscopic aspect, this paper, using multiple linear regression models, has set up the middle-term (1-5 years) prediction of the total quantity of cigarette sales. In addition, practical tests on the relation of population growth and consumption are performed. The research shows that this multiple linear regression model applied to the forecast of the total quantity of cigarette sales can get accurate and credible results and can improve the precision reliability.