“…With the continuous progress of technology and methods, data mining technology will bring about the transformation of research methods in the field of marketing and consumer behavior from deduction to induction [ 33 ]. In solving the research of consumer behavior mining, more researchers began to study customer lifetime value and its components based on machine learning algorithms, such as establishing customer churn prediction models through machine learning methods, such as classification tree [ 34 ], GAM [ 35 , 36 ], SVM [ 37 ], and random forest [ 21 ]. Chen and Fan [ 38 ] used support vector regression to predict dynamic customer lifetime value and used customer behavior variables and controlled variables about multiple promotions to select the best promotion.…”