Customer Segmentation Based on Machine Learning Methods

In recent years, as times change, consumer behavior continues to change rapidly, and their preferences and consumer attitudes change with age and experience. In the generalization of the mass market, it is difficult to identify the needs and desires of customers through various promotional tools. Therefore, customer segmentation can be an option for marketers to offer preferential goods or services to customers. Segmentation can help the company to quickly identify the preferences of the customers and provide them with the desired goods. However, there are significant differences between customers, making it difficult for merchants to segment customers through simple attribute filtering. Fortunately, with the development of machine learning, machine learning-based customer segmentation methods have received a lot of attention from researchers. However, different machine learning methods have different characteristics and there are some differences in commercial applications. Therefore, this paper analyses the principles and performance of the algorithms to provide reference for researchers in related fields. Firstly, this paper introduces several common machines learning methods, including Logistic Regression, Decision Tree, Random Forest and AdaBoost, and then compares the effectiveness of these algorithms through experiments. Finally, this paper looks forward to future research directions.