题名 |
資料探勘在銀行信貸行銷上之應用 |
并列篇名 |
Applying Data Mining to Credit Marketing |
DOI |
10.6338/JDA.201108_6(4).0009 |
作者 |
李御璽(Yue-Shi Lee);顏秀珍(Show-Jane Yen) |
关键词 |
屬性篩選 ; 分群 ; 資料探勘 ; 因素分析 ; 失誤值 ; Attribute Selection ; Clustering ; Data Mining ; Factor Analysis ; Missing Value |
期刊名称 |
Journal of Data Analysis |
卷期/出版年月 |
6卷4期(2011 / 08 / 01) |
页次 |
131 - 157 |
内容语文 |
繁體中文 |
中文摘要 |
借貸是銀行主要業務之一。隨著銀行客戶人數的迅速成長及龐大呆帳的累積,對所有客戶進行信貸行銷的傳統作法,已逐漸不適用。針對可能對信貸有興趣且風險較低的客戶,有效率地進行信貸行銷,才能使銀行節省可觀的成本,同時降低呆帳的發生率。本研究使用某銀行所提供與客戶相關的資料,應用資料探勘(Data Mining)中的分群(Clustering)技術,將客戶分群,並分析各個群組的特性,協助銀行行銷部門,直接針對可能對信貸有興趣且風險較低的群組,進行行銷。本研究首先處理資料集中,有失誤值(Missing Value)的屬性,並從既有的屬性中,新增一些與信貸有關的屬性,再以因素分析(Factor Analysis)法,進行屬性篩選(Attribute Selection),最後以SOM演算法進行資料分群。從分群結果的分析中,本研究找出數個可針對其進行信貸行銷的群組。 |
英文摘要 |
Loan is one of the major businesses in a bank. With the growing amounts of customers and great bad debts, it is not suitable to promote the credit to all customers. To target interested customers in credit with low risk is a feasible way. It can significantly reduce the marketing cost and decrease the occurrence of bad debt in a bank. According to the customer dataset provided by a bank, we use the clustering techniques to group the customers into several clusters and to find the interesting characteristics for each cluster. These analyses can help the sales department to provide the credit to the right customers. This paper firstly deals with missing attribute values in the dataset. Then, we use the original attributes to generate some new attributes related to credit. Before executing SOM clustering algorithm, factor analysis is used to select the important attributes. Based on the clustering results, several clusters can be identified as good promotion groups. |
主题分类 |
基礎與應用科學 >
資訊科學 基礎與應用科學 > 統計 社會科學 > 管理學 |
参考文献 |
|
被引用次数 |