隨著客戶關系管理系統(tǒng)的不斷發(fā)展和應用，使用先進的算法進行客戶分析變得越來越重要。尤其是象銀行這種以客 戶為導向的行業(yè)，客戶分析是十分必要的。當前，支持向量機方法作為一種統(tǒng)計學習理論的分類方法已經(jīng)發(fā)展的比較成熟而 且成功應用到了很多領域。文章解決的主要問題是對銀行的客戶數(shù)據(jù)根據(jù)其屬性對客戶進行分類，為銀行的客戶關系管理系 統(tǒng)提供一種可靠的分類方法。文中主要介紹了銀行的客戶分類學習的過程和結果，如，客戶數(shù)據(jù)清洗，數(shù)據(jù)預處理，SVM 進 行數(shù)據(jù)分類，多類分類處理，客戶屬性選擇等問題。
關鍵詞：客戶關系管理；支持向量機；數(shù)據(jù)挖掘；客戶分類;屬性選擇
Abstract：With the development and application of the customer relationship management, it become more important to analyze the customer using advanced data mining algorithms. And recently Support Vector Machine, a statistical theory method, has been successfully developed and applied to many areas. Support Vector Machine is chose as the data mining method in the analytical Customer Relationship Management system. The main problems are the customer classification, potential customers finding based on bank credit customer data which is mainly about the properties of the customers. The customer data preprocess and data
classification process by using Support Vector Machines, and some results are presented in detail.
Keywords: CRM (customer relationship management); SVM (support vector machine); DM (data mining); customer classification; attribute selection