Customer Segmentation and Anticipation of Consumer Behaviors Based on Machine Learning and CART

Companies should choose competitive markets for their products to maximize the efficiency of their resources. The younger generation has increasing demands for personalized products. The way their needs are met and understanding their consumer behaviors should allow companies to continuously analyze trends within the target group. In the current era of globalization, the Internet, and Big Data, using traditional methods alone may lead to companies selecting the wrong markets, resulting in significant financial and resource losses. Therefore, this article proposes the utilization of machine learning and CART for modeling customer needs to facilitate market segmentation among the younger generation, focusing on young individuals studying in the IT field. Implementing such modeling in practice can contribute to optimizing decision-making, minimizing financial losses, and resource efficiency. The study analyzed the needs of 1149 individuals aged 16-19 in the Wielkopolska region. The developed results provide an explanation of the "high salary" and "low salary" values in the form of a decision tree. Additionally, a spatial distribution map of expected salaries was visualized using the CART model. The extracted rules, which can be explicitly interpreted for each terminal node of the CART model, not only allow for spatial differentiation of the model but, most importantly, enable understanding of the motivations driving the survey respondents. The conducted research demonstrated that to comprehend the motivations of the surveyed individuals, it is crucial to consider several completely different independent variables in the process of modeling the spatial distribution of expected remuneration, including economic parameters. The conducted analyses have shown that machine learning and AI have broad applications in marketing, including customer and market segmentation.