Research on the awakening Algorithm of sleeping members in drugstore

The Awakening of sleeping members is the focus of smart marketing. The cost of losing a sleeping member in existing pharmacies is about 1/4 of the cost of adding a new member. This paper abstractly classifies the problem of whether the sleeping members are easy to be woken up in drugstore into a dichotic problem, and solves the problem that the existing sleeping members' wake-up model is applied to the users of the sleeping members in drugstore and the prediction accuracy of the users' consumption in the store is not high. According to sleep members have rich behavior characteristics, behavior, medical conditions, gender, age, payment, the membership card integral, active level characteristics such as properties, through the analysis of these sleep member attributes in order to distinguish between normal users and sleep, characteristics and the actual scene at the same time, using the method of feature selection to remove redundant features to form a new expression model. On the basis of multiple perspectives, multiple algorithm models were integrated to wake up the sleeping members. The actual store-to-store consumption data of the sleeping member users were used to verify and compare the effectiveness of the algorithm model, and the characteristic expression model with the best discrimination and accuracy was selected. Through the experimental study of the abstract dichotomy algorithm, it is found that the sleeping members with what characteristics are easy to be woken up. The experiment proves that the feature expression model has a high degree of differentiation for the sleeping users, and has a high accuracy in judging the users who are likely to be woken up among the sleeping members of the pharmacy.