IDENTIFICATION OF A POTENTIAL CONSUMER OF E-COMMERCE MARKET PRODUCTS BY GRADIENT BUSTING METHOD

Abstract

E-commerce is an integral part of the developed economy in the country. Small and large businesses can sell their products or services online, meeting the needs of consumers anywhere and anytime. The development of e-commerce is impossible without the knowledge of consumer behavior. Data has become one of the world's most valuable resources due to the rapid digital transformation of global industries. Collecting customer data has become a top priority for businesses. As more and more advanced technologies are developed to collect and analyze customer data, more companies are able to contextualize, retrieve and monetize information from them. Knowing why people buy, companies can grow more effectively in e-commerce and do so strategically, knowing what next steps to take. From consumer behavior to predictive analytics, companies regularly collect, store and analyze large amounts of quantitative and qualitative data about their consumer base on a daily basis. E-commerce services are one of the areas where new data is collected every day. Therefore, it seems necessary to use data analysis methods in this area. User interactions with e-commerce platforms are complex patterns of behavior that, if analyzed, can enable businesses to understand consumer needs. Consumer buying behavior is influenced by many factors, and different consumer demands lead to large differences in consumer buying behavior. To predict consumer buying behavior it is necessary to determine the hidden characteristics of data in the array of information left by users on the e-commerce platform, and then to determine the desire of future users to buy on the e-commerce platform. The article considers an example of application of the gradient boosting algorithm for consumer identification in the e-commerce market.