資料探勘分類技術於游泳會員流失區別模型之研究
No Thumbnail Available
Date
2004
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
本研究旨在暸解台灣師大本部游泳會員之組成結構，運用資料探勘中的鑑別分析、類神經網路、多元適應性雲形迴歸以及整合類神經網路與多元適應性雲形迴歸等分類技術建構台灣師大游泳會員流失區別模型，並瞭解會員流失的重要特徵。
台灣師大本部游泳會員資料經整理後，共2,707筆，在剔除內容不合之資料後，共計2,380筆。本研究結果如下：
一、會員組成結構如下：(一)男(49.87%)、女(50.13%)會員幾近相等；(二)會員類型以自由會員(55.63%)稍多；(三)大部分會員居住在大安中正兩區(79.12%)；(四)會員平均年齡為32.15歲且年齡的分佈平均沒有特別集中的現象；(五)會員平均會齡為0.69年且大量集中在二年以下(93.66%)；(六)沒有折扣會員(95.80%)佔大多數；(七)繳費金額以4,500元(42.86%)最多；(八)購買季節以夏季(49.66 %)居多；(九)會員使用時段以不受限制的任何時段(56.64%)最多。
二、整合類神經網路與多元適應性雲形迴歸分析模型的整體分類績效最高，為84.03 %。整合模式成功地建構台灣師大本部游泳會員流失區別模型。
三、台灣師大本部游泳會員流失的重要特徵為會齡在1年以下、繳費金額為2,500元、購買季節在夏季的一般類型會員。
有鑑於此，台灣師大游泳池管理單位可就分析的結果轉為會員維繫方案，達到降低會員流失的目的。
The purpose of this study was to provide the realization of member constitutive structure in National Taiwan Normal University(NTNU) main campus swimming pool, and the member churn model of NTNU main campus swimming pool was established by data mining classification technology including discriminant analysis, artificial neural networks, multivariate adaptive regression splines, combining artificial neural networks and multivariate adaptive regression splines, and the important characteristics of member churn were also emphasized. After reorganizing the data of NTNU main campus swimming members, 2,707 records were chosen as original data. After deleting the unreasonable data, totally, there were 2,380 records discussed in this study. The research results were as follows: 1. The constitutive structure of members indicated that: (1) Male members (49.87%) and female members (50.13%) were almost even; (2) Free type members were slightly more (55.63%); (3) The most members were the residents of Da-an and Zhongzheng district (79.12%); (4) The average age of members was 32.15-year- old and distributed evenly without specially centralized phenomenon; (5) The average participation period of members was 0.69 year and mostly centralized on 2 years below (93.66%); (6) No discount members were at major (95.80%); (7) 4,500 NTdollars of member fee was at most (42.86%); (8) Enrollment membership in summer season was at major (49.66%); (9) Any time interval without time limitation was at most (56.64%). 2. Combining artificial neural networks and multivariate adaptive regression splines implied the best whole correct classification rate, 84.03%. The integrated approach successfully constructs the member churn model of NTNU main campus swimming pool. 3. The important characteristics of NTNU main campus swimming member churn were participation period below 1 year, the member fee of 2,500 NT dollars, membership enrollment in summer, and general members. Base on this, NTNU swimming pool administration deparment may transfer the results to the plan of maintaining members in order to decrease member churn.
The purpose of this study was to provide the realization of member constitutive structure in National Taiwan Normal University(NTNU) main campus swimming pool, and the member churn model of NTNU main campus swimming pool was established by data mining classification technology including discriminant analysis, artificial neural networks, multivariate adaptive regression splines, combining artificial neural networks and multivariate adaptive regression splines, and the important characteristics of member churn were also emphasized. After reorganizing the data of NTNU main campus swimming members, 2,707 records were chosen as original data. After deleting the unreasonable data, totally, there were 2,380 records discussed in this study. The research results were as follows: 1. The constitutive structure of members indicated that: (1) Male members (49.87%) and female members (50.13%) were almost even; (2) Free type members were slightly more (55.63%); (3) The most members were the residents of Da-an and Zhongzheng district (79.12%); (4) The average age of members was 32.15-year- old and distributed evenly without specially centralized phenomenon; (5) The average participation period of members was 0.69 year and mostly centralized on 2 years below (93.66%); (6) No discount members were at major (95.80%); (7) 4,500 NTdollars of member fee was at most (42.86%); (8) Enrollment membership in summer season was at major (49.66%); (9) Any time interval without time limitation was at most (56.64%). 2. Combining artificial neural networks and multivariate adaptive regression splines implied the best whole correct classification rate, 84.03%. The integrated approach successfully constructs the member churn model of NTNU main campus swimming pool. 3. The important characteristics of NTNU main campus swimming member churn were participation period below 1 year, the member fee of 2,500 NT dollars, membership enrollment in summer, and general members. Base on this, NTNU swimming pool administration deparment may transfer the results to the plan of maintaining members in order to decrease member churn.
Description
Keywords
資料探勘, 會員流失, 鑑別分析, 類神經網路, 多元適應性雲形迴歸, Data Mining, Member Churn, Discriminant Analysis, Artificial Neural Networks, Multivariate Adaptive Regression Splines