Research on financial consumer behavior based on deep Learning

At present, the judgment of users' financial literacy is mainly based on users' default records and loan information, but the research on reflecting users' financial literacy through users' consumption behavior is limited. Based on the transaction data of an internet financial company, a fusion model is established for a variety of machine learning algorithms, and the relationship between the consumer behavior and the credit risk of the user is established. At the same time, based on the characteristics of artificial feature selection and information gain of tree model, the main consumption behaviors that affect users' credit risk are obtained, which provides a new idea for judging users' financial literacy through studying and judging users' consumption behaviors.