Analysis method of cigarette competing products based on LASSO regression

Analysis of competing cigarette products is crucial to improving product quality and brand popularity. Additionally, it is a vital part of tobacco industry research content. Due to traditional analysis methods relying on manual experience to identify cigarette competitors, a novel analysis of competing cigarette products based on LASSO regression is proposed in the paper. We first divide the market into several commercial district grids. We obtain the consumption population attributes within each district and the sales data for all cigarettes in circulation. Secondly, we construct several LASSO regression models for different cigarettes and obtain the significant driving factors of each cigarette. Finally, analysing and comparing the significant driving factors of different cigarettes allows us to judge the existence of competitive relationships between them. This will enable us to identify cigarette competitors. Taking 235 cigarette varieties in the Liuzhou market as an example, the analysis shows that the proposed method can effectively screen the main consumer groups of different cigarettes and identify cigarette competitors, providing technical support for tobacco companies to enhance their brand competitiveness.