The study of customer classification based on self-organizing feature map neural network

Under current highly competitive market environment, an increasing number of enterprises begin to value the operation mode that changes from the traditional passive-service mode to active-service mode which provides different customers with different services, and effective customer classification is the premise and basis to implement this mode. Through constructing a SOFM model based on RFM indexes, this paper conducted a customer classification study in retail industry, and the simulation experiments using MATLAB software verified the feasibility and validity of this model. This study hopes to optimize retail enterprises´ resources allocation efficiency and promote economic benefits by making personalized marketing and active-service strategies.