Research on the Development of Urban E-commerce Based on Functional Data

Abstract

With the advancement of the digital economy, the volume of data generated daily is growing exponentially. Traditional statistical analysis methods struggle to handle such large amounts of data. As a significant portion of the collected data is presented in time series, these data demonstrate certain characteristics of functional data as the data volume increases. By expanding these data into functional data using B-spline basis functions, they can be transformed into infinite-dimensional data with function properties. This study utilizes the transaction data of an e-commerce platform throughout a day, applies functional principal component analysis, and analyzes the development of e-commerce in major cities across the country. Future marketing recommendations for these cities are also proposed.