本文選題：支持向量機 + 客戶關系管理　； 參考：《南昌大學》2015年碩士論文

【摘要】：目前,大量外資銀行涌入我國,給我國銀行帶來了強大的沖擊力。這就需要我國重視銀行的信息化建設,其中,對于銀行客戶數(shù)據(jù)的管理顯得尤為重要。以客戶為中心是當今銀行研究的熱點之一,良好的客戶關系管理能為銀行帶來巨大的利益。但是,海量的客戶數(shù)據(jù)僅僅靠人工的方式進行管理已經(jīng)顯得力不從心。數(shù)據(jù)挖掘技術很好的解決了這一問題。通過數(shù)據(jù)挖掘技術預測客戶行為,支持銀行做出決策,為客戶提供不同的服務方式和產(chǎn)品。支持向量機是數(shù)據(jù)挖掘的一種新方法,以其結構風險最小化、解決維數(shù)災等特征而成為研究熱點。支持向量機可以很好的將海量數(shù)據(jù)進行分類,是大數(shù)據(jù)時代很好的機器學習方法。本文中我們對支持向量機和客戶關系管理進行了理論研究,同時利用支持向量機算法對銀行客戶關系管理中的客戶數(shù)據(jù)進行了細分操作。支持向量機現(xiàn)階段主要應用于二分類問題,在多分類方面應用較少,本文中將支持向量機應用于銀行客戶細分的多分類問題,是一個創(chuàng)新點。文中主要利用支持向量機SVM對銀行客戶關系管理中的銀行數(shù)據(jù)進行分類預測,從而驗證SVM在多分類問題中的準確率如何,進而協(xié)助銀行對未知分類的客戶進行分類操作,同時證明支持向量機在多分類問題中也有很好前景。
[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.
【學位授予單位】：南昌大學
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：F832.2;TP18

【參考文獻】

相關期刊論文 前1條

1 馮振華;楊潔明;;SVM回歸的參數(shù)選擇探討[J];機械工程與自動化;2007年03期

相關碩士學位論文 前1條

1 張永新;基于SVM的人臉檢測算法研究[D];西北大學;2009年

，

本文編號：1884456