基于群体行为模式特征的银行客户价值划分模型研究

发布时间：2018-03-28 05:10

本文选题：社区群体行为模式特征　切入点：客户价值划分　出处：《天津大学》2016年硕士论文

【摘要】：社区群体行为模式特征能够反应群体行为在时间特性和空间特性上存在的规律。使用社区属性、社区群体行为模式特征和银行领域特征进行相关性分析,获取银行领域指标,构建基于群体行为模式特征的银行客户价值划分模型,能够帮助银行更好地进行客户价值划分。本文的工作内容如下:第一,提出了基于群体行为模式特征的银行客户价值划分模型,模型的建立过程为特征集选取,特征集包括了社区属性、社区群体行为模式特征和银行领域特征,并在此基础上进行相关性分析,以及客户价值聚类模型构建。第二,提出了银行客户价值划分平台设计,分别是平台的总体设计和模块设计。总体设计包含了框架设计、流程图设计和接口设计;模块设计包括社区数据获取模块设计、社区群体行为模式特征计算模块设计、银行领域数据获取模块设计、相关性分析模块设计和银行客户价值划分模型模块设计。第三,提出了银行客户价值划分模型的实验验证与评估。实验验证包括了数据集获取、相关性分析实验和银行客户价值划分模型构建验证。其中,相关性分析实验包括了社区属性与群体行为模式特征的相关性分析、社区属性与银行领域特征的相关性分析和社区群体行为模式特征与银行领域特征的相关性分析。本文构建的基于群体行为模式特征的银行客户价值划分模型,对于银行的客户价值划分具有一定的指导意义。同时,本文提出的基于银行客户价值划分的平台设计为后续的开发起到了推动作用。
[Abstract]:The characteristics of community behavior patterns can reflect the temporal and spatial characteristics of community behavior. Using community attributes, community behavior patterns and banking domain characteristics, the correlation analysis is carried out to obtain the banking domain indicators. Building a bank customer value division model based on the characteristics of group behavior patterns can help banks to better divide customer value. The work of this paper is as follows: first, This paper proposes a bank customer value partition model based on the characteristics of group behavior pattern. The process of establishing the model is the selection of feature set, which includes community attribute, community group behavior pattern feature and banking domain feature. And on the basis of this correlation analysis, and customer value clustering model construction. Second, the bank customer value division platform design, respectively, the overall design of the platform and module design, the overall design includes the framework design, Flow chart design and interface design, module design includes community data acquisition module design, community behavior pattern calculation module design, bank domain data acquisition module design, The correlation analysis module design and the bank customer value division model module design. Third, proposed the bank customer value division model experimental verification and evaluation, the experimental verification includes the data set acquisition, The correlation analysis experiment and the bank customer value division model construction validation, among them, the correlation analysis experiment includes the community attribute and the group behavior pattern characteristic correlation analysis. The correlation analysis between community attributes and banking domain characteristics and the correlation analysis between community behavior pattern characteristics and banking domain characteristics. This paper constructs a bank customer value division model based on the characteristics of community behavior patterns. At the same time, the design of the platform based on the bank customer value division plays an important role in the subsequent development.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F832;TP311.13

【相似文献】