Research on Mobile Internet User Behavior Preference Based on Support Vector Machine

Abstract

With the rapid development of the mobile Internet, mobile Internet users grow rapidly and the preference is more diverse, and the user's behavior is also showing a new feature. In this study, we use qualitative research and quantitative research, combined with empirical data and simulation analysis and the combination of traditional statistical analysis and modern data mining methods, to study the behavior preference of mobile Internet. Besides, we use the discrete choice model to build the binaryLogit model using SAS tools, and analyze the selected variables through data processing. Based on the significant influence factors, the behavior preference of mobile Internet users is classified by using support vector machine. The C-SVM binary classification machine is used as kernel function, and the parameters of the model and kernel function are optimized by cross validation. In this research, we introduce the support vector machine theory into the research field of the mobile Internet user behavior effectively, solve the problem of small and medium-sized samples effectively, and provide new ideas and methods for the research of mobile Internet user behavior.