【摘要】： 經(jīng)濟(jì)全球化以及電子商務(wù)的實(shí)施,使商業(yè)銀行面臨更加激烈的競(jìng)爭(zhēng),特別是由于對(duì)客戶資源的競(jìng)爭(zhēng)全球化以及爭(zhēng)奪高價(jià)值客戶競(jìng)爭(zhēng)的日趨激烈,使得客戶擾動(dòng)加劇、客戶流失嚴(yán)重、客戶獲得成本增加,銀行經(jīng)營(yíng)風(fēng)險(xiǎn)增加、競(jìng)爭(zhēng)力受到巨大的沖擊。研究發(fā)現(xiàn)在銀行行業(yè),客戶保持是CRM策略成功的關(guān)鍵,只有當(dāng)客戶隨著時(shí)間得到保持,這一策略才是有利可圖的,并且成功的客戶保持能夠降低銀行尋求新的、具有潛在風(fēng)險(xiǎn)客戶的需求,并使銀行將注意力集中在建立關(guān)系和滿足現(xiàn)有客戶的需求上。所以面對(duì)當(dāng)前的市場(chǎng)狀況,商業(yè)銀行必須在發(fā)展新客戶的同時(shí),著手進(jìn)行客戶保持的研究。商業(yè)銀行維系現(xiàn)有客戶可以增加資金的匯聚量,同時(shí)節(jié)省誘導(dǎo)客戶進(jìn)入銀行所必需的廣告和介紹成本,從而產(chǎn)生更多的現(xiàn)金流和利潤(rùn)。銀行客戶保持的成功主要依靠對(duì)銀行客戶流失情況的分析和評(píng)估,從而提前預(yù)知某些客戶是否有流失的可能性,進(jìn)而采取市場(chǎng)策略。而空前巨大的個(gè)體水平客戶數(shù)據(jù)量使得銀行的數(shù)據(jù)庫(kù)變得更加巨大和復(fù)雜,數(shù)據(jù)挖掘技術(shù)能夠勝任對(duì)海量數(shù)據(jù)的處理,必將在銀行行業(yè)的客戶流失分析中發(fā)揮巨大的作用,從海量的普通業(yè)務(wù)數(shù)據(jù)中發(fā)掘出關(guān)于客戶流失的關(guān)鍵信息,幫助銀行留住最寶貴的資源——客戶。 本文正是在這樣的背景下,運(yùn)用市場(chǎng)營(yíng)銷、管理決策理論與方法、數(shù)據(jù)挖掘技術(shù)和統(tǒng)計(jì)技術(shù),圍繞具有廣泛的實(shí)際背景和發(fā)展前景的商業(yè)銀行客戶關(guān)系管理核心部分客戶流失問(wèn)題進(jìn)行了系統(tǒng)的研究。首先在考察和分析其它相關(guān)研究使用的變量基礎(chǔ)上,得出與客戶流失密切相關(guān)的因素:在銀行服務(wù)時(shí)間長(zhǎng)度,年齡,與銀行接觸的主要渠道,是否購(gòu)買銀行的某些產(chǎn)品,擁有銀行各種業(yè)務(wù)的數(shù)量等13個(gè)與銀行客戶流失密切相關(guān)的靜態(tài)客戶資料;同時(shí),引入了時(shí)間序列的因素,即考察期前一年銀行客戶的各種交易行為,最終把這兩部分因素都做為客戶流失模型的輸入變量,最終輸入模型的解釋變量達(dá)到了200多個(gè)。使用Weka和SAS Enterprise Miner兩種數(shù)據(jù)挖掘軟件分別建立了某商業(yè)銀行客戶流失的決策樹預(yù)測(cè)模型和Logistic回歸預(yù)測(cè)模型,最后對(duì)所建立的客戶流失預(yù)測(cè)模型的預(yù)測(cè)效果進(jìn)行了比較和分析,識(shí)別了商業(yè)銀行將要流失客戶的特征。研究結(jié)果對(duì)于商業(yè)銀行設(shè)計(jì)銀行客戶保持規(guī)劃,維系有價(jià)值客戶,提高商業(yè)銀行基于事實(shí)的決策制定能力,通過(guò)與有價(jià)值的客戶保持長(zhǎng)期穩(wěn)定的關(guān)系增加客戶對(duì)銀行利潤(rùn)的貢獻(xiàn),和幫助銀行獲得真正的競(jìng)爭(zhēng)優(yōu)勢(shì)具有十分重要的理論價(jià)值和現(xiàn)實(shí)意義。
[Abstract]:Economic globalization and the implementation of e-commerce make commercial banks face more fierce competition, especially due to the globalization of competition for customer resources and the increasingly fierce competition for high-value customers, which aggravates the customer disturbance. The loss of customers is serious, the cost of customer acquisition is increased, the risk of bank operation is increased, and the competitiveness is greatly impacted. The study found that in the banking industry, customer retention is key to the success of the CRM strategy, which is profitable only when the customer is maintained over time, and that successful customer retention lowers the bank's search for new ones. Demand from customers with potential risks and focus on building relationships and meeting the needs of existing customers. Therefore, in the face of the current market conditions, commercial banks must develop new customers, at the same time, proceed with customer maintenance research. Maintaining existing customers by commercial banks can increase the amount of capital gathered while saving the advertising and introduction costs necessary to induce customers to enter the bank, thereby generating more cash flow and profits. The success of bank customers depends mainly on the analysis and evaluation of bank customer turnover, so as to predict the possibility of some customers losing ahead of time, and then adopt the market strategy. And the unprecedented individual level of customer data makes the database of the bank more huge and complex, data mining technology can be competent to deal with massive data, will play a huge role in the analysis of customer churn in the banking industry. Discover key information about customer churn from the vast amount of ordinary business data to help banks retain the most valuable resource-customers. Under this background, this paper applies marketing, management decision theory and method, data mining technology and statistics technology, and makes use of the theory and method of marketing, management decision-making, data mining and statistics. This paper makes a systematic study on customer churn, which is the core part of customer relationship management in commercial banks, which has a wide practical background and development prospects. First of all, on the basis of investigating and analyzing the variables used in other related studies, the factors closely related to customer loss are obtained: the length of bank service, the age, the main channel of contact with the bank, and whether to purchase certain products of the bank. Has the bank various business quantity and so on the 13 static customer information which is closely related to the bank customer loss; At the same time, the factors of time series, that is, the transaction behavior of bank customers in the year before the survey period, are introduced. At last, these two factors are regarded as input variables of customer churn model, and the explanatory variables of the input model reach more than 200. Two kinds of data mining software, Weka and SAS Enterprise Miner, are used to establish the decision tree forecasting model and Logistic regression forecasting model of customer churn in a commercial bank, respectively. Finally, the forecasting effect of the established customer churn prediction model is compared and analyzed. Identify the characteristics that commercial banks are about to lose customers. The research results maintain the planning of commercial bank design bank customers, maintain valuable customers, and improve the ability of commercial banks to make decision-making based on facts. It is of great theoretical value and practical significance to increase customers' contribution to bank profits by maintaining long-term stable relationship with valuable customers and to help banks gain real competitive advantage.
【學(xué)位授予單位】：重慶大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2008
【分類號(hào)】：F830.4;F224

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 汪楠;成鷹;崔婀娜;;數(shù)據(jù)挖掘技術(shù)怎么用[J];經(jīng)營(yíng)與管理;2007年07期

2 余云飛;;數(shù)據(jù)挖掘與科學(xué)決策[J];決策與信息;2008年09期

3 肖娟;葉楓;;利用數(shù)據(jù)挖掘管理客戶關(guān)系[J];經(jīng)濟(jì)管理;2003年01期

4 楊華;;淺析數(shù)據(jù)挖掘在企業(yè)信息化管理中應(yīng)用現(xiàn)狀和對(duì)策[J];江西金融職工大學(xué)學(xué)報(bào);2007年05期

5 尚銳;陳超;;數(shù)據(jù)挖掘：計(jì)算機(jī)審計(jì)的新思路[J];審計(jì)月刊;2006年22期

6 商錦博;;探索數(shù)據(jù)挖掘在保險(xiǎn)公司中的應(yīng)用[J];商場(chǎng)現(xiàn)代化;2007年19期

7 張永賓;;DM在分析客戶忠誠(chéng)度的應(yīng)用[J];商場(chǎng)現(xiàn)代化;2008年34期

8 劉忠寶;尤忠彬;;數(shù)據(jù)挖掘在信用卡客戶營(yíng)銷中的運(yùn)用[J];中國(guó)信用卡;2008年04期

9 宋艷;張炎欣;;數(shù)據(jù)挖掘技術(shù)在CRM中的應(yīng)用[J];企業(yè)導(dǎo)報(bào);2009年04期

10 丁文捷;;基于量體數(shù)據(jù)挖掘的服裝投產(chǎn)號(hào)型決策[J];統(tǒng)計(jì)與決策;2007年17期

相關(guān)會(huì)議論文 前10條

1 馬杰;胡海峰;馬玉書;;基于人工神經(jīng)網(wǎng)絡(luò)的數(shù)據(jù)挖掘工具[A];第十七屆全國(guó)數(shù)據(jù)庫(kù)學(xué)術(shù)會(huì)議論文集（技術(shù)報(bào)告篇）[C];2000年

2 馬杰;劉月田;胡海峰;馬玉書;;具有多種神經(jīng)網(wǎng)絡(luò)模型的數(shù)據(jù)挖掘工具[A];第十九屆全國(guó)數(shù)據(jù)庫(kù)學(xué)術(shù)會(huì)議論文集（技術(shù)報(bào)告篇）[C];2002年

3 楊慶躍;楊冬青;唐世渭;王騰蛟;;移動(dòng)通信客戶流失分析中的數(shù)據(jù)預(yù)處理方法研究[A];第二十二屆中國(guó)數(shù)據(jù)庫(kù)學(xué)術(shù)會(huì)議論文集（技術(shù)報(bào)告篇）[C];2005年

4 邵華;萬(wàn)家華;王劍虎;何耀東;張霞;;一個(gè)以用戶為中心的數(shù)據(jù)挖掘工具:Open Miner[A];第十七屆全國(guó)數(shù)據(jù)庫(kù)學(xué)術(shù)會(huì)議論文集（研究報(bào)告篇）[C];2000年

5 徐進(jìn);;數(shù)據(jù)挖掘在客戶關(guān)系管理中的應(yīng)用[A];2007中國(guó)科協(xié)年會(huì)——通信與信息發(fā)展高層論壇論文集[C];2007年

6 陳震;陳維默;;淺談數(shù)據(jù)挖掘技術(shù)[A];福建省科協(xié)第五屆學(xué)術(shù)年會(huì)數(shù)字化制造及其它先進(jìn)制造技術(shù)專題學(xué)術(shù)年會(huì)論文集[C];2005年

7 段軍曉;;數(shù)據(jù)挖掘技術(shù)在民航快遞市場(chǎng)營(yíng)銷中的應(yīng)用[A];第六屆中國(guó)科學(xué)家論壇論文匯編[C];2007年

8 連建勇;;基于數(shù)據(jù)挖掘的電信客戶流失分析[A];2007中國(guó)科協(xié)年會(huì)——通信與信息發(fā)展高層論壇論文集[C];2007年

9 崔海莉;袁兆山;;客戶關(guān)系管理中的關(guān)聯(lián)規(guī)則挖掘技術(shù)[A];計(jì)算機(jī)技術(shù)與應(yīng)用進(jìn)展——全國(guó)第17屆計(jì)算機(jī)科學(xué)與技術(shù)應(yīng)用（CACIS）學(xué)術(shù)會(huì)議論文集（上冊(cè)）[C];2006年

10 李忍;孫晶;劉令君;;一個(gè)體育比賽技戰(zhàn)術(shù)分析數(shù)據(jù)挖掘工具的設(shè)計(jì)與實(shí)現(xiàn)[A];2007通信理論與技術(shù)新發(fā)展——第十二屆全國(guó)青年通信學(xué)術(shù)會(huì)議論文集（上冊(cè)）[C];2007年

相關(guān)重要報(bào)紙文章 前10條

1 劉慶;關(guān)于數(shù)據(jù)挖掘工具的討論[N];網(wǎng)絡(luò)世界;2006年

2 張少榮;除了技術(shù) 還需要什么[N];金融時(shí)報(bào);2006年

3 毛江華;6030億元 客戶流失威脅中國(guó)企業(yè)發(fā)展[N];計(jì)算機(jī)世界;2008年

4 汪雯;如何進(jìn)行數(shù)據(jù)挖掘[N];金融時(shí)報(bào);2006年

5 袁聞;數(shù)據(jù)挖掘和在線分析處理[N];金融時(shí)報(bào);2006年

6 記者　 王璐;數(shù)據(jù)挖掘 提升銀行核心競(jìng)爭(zhēng)力[N];金融時(shí)報(bào);2006年

7 陳金波;數(shù)據(jù)挖掘：提升電信CRM水平的助推器[N];人民郵電;2007年

8 梅靜彥;數(shù)據(jù)挖掘技術(shù)在美國(guó)銀行的應(yīng)用[N];金融時(shí)報(bào);2006年

9 NCR(中國(guó))有限公司數(shù)據(jù)倉(cāng)庫(kù)事業(yè)部 張新宇;開放標(biāo)準(zhǔn)提升數(shù)據(jù)挖掘效率[N];通信產(chǎn)業(yè)報(bào);2006年

10 高蕓;數(shù)據(jù)挖掘技術(shù)：整合數(shù)據(jù)庫(kù)信息的利器[N];中國(guó)城鄉(xiāng)金融報(bào);2005年

相關(guān)博士學(xué)位論文 前10條

1 艾丹祥;基于數(shù)據(jù)挖掘的客戶智能研究[D];武漢大學(xué);2007年

2 徐雪琪;基于統(tǒng)計(jì)視角的數(shù)據(jù)挖掘研究[D];浙江工商大學(xué);2007年

3 晏峻峰;基于數(shù)據(jù)挖掘技術(shù)的證素辨證方法研究[D];湖南中醫(yī)藥大學(xué);2007年

4 李榮;生物信息數(shù)據(jù)挖掘若干關(guān)鍵問(wèn)題研究與應(yīng)用[D];復(fù)旦大學(xué);2004年

5 劉君強(qiáng);海量數(shù)據(jù)挖掘技術(shù)研究[D];浙江大學(xué);2003年

6 史e，

本文編號(hào)：2440406