Precision Marketing Scheme based on Integrating Spatio-temporal Data Clustering and Neural Network

Abstract

Precision marketing can effectively predict the user’s willingness to consume, with nearly one trillion of market space. The current precision marketing strategy still has remarkable problems such as the low correlation between data model and space; the marketing initial data redundancy and so on. Aiming at the low correlation between the model and space and the initial data redundancy in the marketing, this paper proposes a precise marketing scheme combining the spatiotemporal data clustering and the neural network. Through the establishment of multi - dimensional space - time large data model, effectively enhance the geographic data model, spatio - temporal data model and multi - dimensional data superposition. Design the large-scale data normalization clustering scheme, construct the normalized classification Euclidean distance and the adaptive precision marketing objective function, and realize the K-Means data clustering of the initial data redundancy. Through the neural network based spatial and temporal data precision marketing decision, design data input layer; deal with hidden layer and intelligent output layer; provide an effective program for accurate marketing.