聚类分析在H银行客户分类中的应用

发布时间：2018-01-20 16:17

本文关键词： 数据挖掘聚类分析数据化运营营销策略　出处：《华南理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：聚类分析是数据挖掘的重要功能之一,近年来在该领域的研究取得了长足的发展。聚类分析方法所涉及的领域几乎遍及人工智能的方方面面,在各行各业以信息分析为基础的决策支持系统活动中扮演越来越重要的角色。它在电子商务、图像处理,模式识别、文本分类等领域有广泛的应用。本文对聚类分析算法在客户分类中的应用进行了深入的研究,主要研究工作如下:1)深入研究聚类分析技术,对聚类分析中的各种算法进行了详细的分析。2)重点研究聚类分析算法中应用比较广泛的系统聚类算法。对过比较系统聚类算法与其它聚类算法的优缺点,分析各个方法的适用性,以及系统聚类在客户分类中的适用性。3)在深入研究了系统聚类算法的基础上,对H银行网上转账系统的客户进行分类。并对分类过程中所涉及的指标参数变量选取、距离的度量、数据市集的建立以及数据预处理等方面进行了详细的论述。4)对使用系统聚类算法进行客户分类后的结果叙述验证,最后结果实践证明该聚类算法在客户分类上的有效性。通过对分类后筛选出来的有价值客户群体进行定向营销,付费客户数量的提升率比起在测试运营阶段有所提升。本文将数据挖掘中的聚类分析引入到银行的营销策略分析中来,为市场的营销战略与策略提供了非常科学的参考体系,也为各类企业在数据化运营的方向上提供了非常有价值的实践。
[Abstract]:Clustering analysis is one of the important functions of data mining. In recent years, great progress has been made in the research in this field. Clustering analysis involves almost every aspect of artificial intelligence. It plays an increasingly important role in the activities of decision support systems based on information analysis in various industries. It plays an increasingly important role in electronic commerce, image processing, and pattern recognition. Text classification and other fields have a wide range of applications. In this paper, the application of clustering analysis algorithm in customer classification has been deeply studied, the main research work is as follows: 1) deeply study clustering analysis technology. This paper makes a detailed analysis of all kinds of algorithms in clustering analysis. (2) focusing on the system clustering algorithm which is widely used in clustering analysis algorithm, the advantages and disadvantages of over-comparison system clustering algorithm and other clustering algorithms are discussed. Analysis of the applicability of each method, and the applicability of system clustering in customer classification. 3) on the basis of in-depth study of the system clustering algorithm. The customer of H bank online transfer system is classified, and the parameter variables are selected and the distance is measured in the process of classification. The establishment of the data market and data preprocessing are discussed in detail. 4) the results of customer classification using the system clustering algorithm are described and verified. Finally, the practice proves the effectiveness of the clustering algorithm in customer classification. Through the classification of valuable customer groups selected out of the targeted marketing. The increase rate of the number of paying customers is higher than that in the test operation phase. This paper introduces the clustering analysis of data mining into the marketing strategy analysis of banks. It provides a very scientific reference system for marketing strategy and strategy, and also provides a very valuable practice for all kinds of enterprises in the direction of data operation.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP311.13

【参考文献】