基于支持向量机SVM的银行客户关系管理研究

发布时间：2018-05-13 19:17

本文选题：支持向量机 + 客户关系管理　；参考：《南昌大学》2015年硕士论文

【摘要】：目前,大量外资银行涌入我国,给我国银行带来了强大的冲击力。这就需要我国重视银行的信息化建设,其中,对于银行客户数据的管理显得尤为重要。以客户为中心是当今银行研究的热点之一,良好的客户关系管理能为银行带来巨大的利益。但是,海量的客户数据仅仅靠人工的方式进行管理已经显得力不从心。数据挖掘技术很好的解决了这一问题。通过数据挖掘技术预测客户行为,支持银行做出决策,为客户提供不同的服务方式和产品。支持向量机是数据挖掘的一种新方法,以其结构风险最小化、解决维数灾等特征而成为研究热点。支持向量机可以很好的将海量数据进行分类,是大数据时代很好的机器学习方法。本文中我们对支持向量机和客户关系管理进行了理论研究,同时利用支持向量机算法对银行客户关系管理中的客户数据进行了细分操作。支持向量机现阶段主要应用于二分类问题,在多分类方面应用较少,本文中将支持向量机应用于银行客户细分的多分类问题,是一个创新点。文中主要利用支持向量机SVM对银行客户关系管理中的银行数据进行分类预测,从而验证SVM在多分类问题中的准确率如何,进而协助银行对未知分类的客户进行分类操作,同时证明支持向量机在多分类问题中也有很好前景。
[Abstract]:At present, a large number of foreign banks pour into our country, which brings a powerful impact to our banks. This requires our country to attach importance to the construction of bank information, among which, the management of bank customer data is particularly important. Taking customer as the center is one of the hot topics in the banking research nowadays. Good customer relationship management can bring huge benefits to the bank. However, massive customer data only rely on manual management has become inadequate. Data mining technology solves this problem very well. Data mining technology is used to predict customer behavior, to support banks to make decisions, and to provide customers with different service modes and products. Support vector machine (SVM) is a new method of data mining, which has become a research hotspot for its structural risk minimization and dimensionality disaster resolution. Support vector machine (SVM) is a good machine learning method in big data era. In this paper, we study the support vector machine and customer relationship management, and use the support vector machine algorithm to subdivide the customer data in bank customer relationship management. Support vector machine (SVM) is mainly applied to the two-classification problem at present, but it is seldom used in multi-classification. In this paper, it is an innovation point to apply SVM to the multi-classification problem of bank customer segmentation. In this paper, support vector machine (SVM) is used to classify and predict bank data in bank customer relationship management (CRM), so as to verify the accuracy of SVM in multi-classification problems. At the same time, it is proved that support vector machine has a good prospect in multi-classification problems.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F832.2;TP18

【参考文献】