Research on User’s Discrimination Algorithm of Power Substitution Based on Logistic Regression Model

State Grid exist the problems of low efficiency and integrating information in the promotion of power substitution, big data technology has the characteristics of high efficiency, high speed, convenient in information processing, can realize the purposes of accurate positioning, differentiated marketing. Based on the analysis of the current State Grid users, the paper summarizes the users’ characteristics of clustering according to the industry and equipment through the "industry field user" three layer structure analysis method, combined with the exploration of the current network of existing data and finishing the digging energy, electricity characteristics, constructing power substitution of user discrimination model, so as to enhance State Grid’s business efficiency and economic benefits. Keywords—Big data, power substitution, User’s discrimination Logistic regression