Bayesiane Ntwork Of Pareto-Inprovement And An Empirical Study Of Marketing

Customer life-cycle theory is theory of the enterprise value of life for the customer analysis. It seperates the links between customers and enterprises into different stages through the statistics and computer knowledge. The characteristics of each stage to take a different marketing method and contact, thereby increasing corporate profits.Data mining is a set of cross-technology of statistics, artificial intelligence,the main task of the data mining is extracting, potentially useful information and knowledge from a large number of, incomplete,noising,vague and random data. As an important component of data mining, classification techniques will be adopted by the knowledge that already exists to guide our learning, while the purpose of learning is building the best model of these data with the knowledge which we have between known deviation of the smallest, which is actually an optimization process. By optimizing, such data should fall into the known knowledge.The main task of this paper describes the customer life-cycle theory of marketing and data mining theory and the status of research and development at home and abroad, introduce the classification techniques of decision tree, neural network and highlights the principle of Bayesian classification techniques and building of model. This article integrate customer life-cycle theory and the theory of data mining,trying to introduce Pareto model of customer life-cycle theory of marketing to improve Bayesian network structure learning algorithm, and its methods will be used in an empirical study of data mining. We are trying to solve a series of data mining theoretical question of markeing,eg, discreting data, and trying to put inprovement of accuracy to evaluate as the evaluation criteria used in marketing of accuracy.Through comparison of the Bayesian network of Pareto-improvement and decision tree and neural networks,showing the advantages of Bayesian theory and the combination of Bayesian networks and marketing will be development direction of data mining. We also expect that the use of Bayesian networks of Pareto-improvement for laying the theoretical foundation of marketing, and thus to guide corporate marketing practices.