Research on Brand Type Decision in Regional Passenger Vehicle Additional Purchase and Trade-in Market Based on CART Decision Tree Algorithm

As passenger vehicle market entered the era of stock competition, seizing opportunities in additional-purchase and trade-in (APT) market is of significant importance for the growth of different brand-type OEMs. This study aims at investigating the decision-making of users in regional APT market regarding brand types and abstracting quantitative models. Data mining and machine learning are used to train CART decision tree models, and the probability of users in different regions and with different characteristics switching to luxury, joint venture, or domestic brands under different car ownerships and repurchase choices are quantified. Common and individual characteristics of brand type decisions among users in different regions are summarised, which provides finer grained and quantifiable references for different types of brands to plumb regional demand for APT market and deploy product layout.