高價值客戶的辨識與購買行為預測

機構典藏 > College of Business and Management > Graduate Institute & Department of Statistics > Thesis > Item 987654321/111095

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/111095

Title:	高價值客戶的辨識與購買行為預測
Other Titles:	Identification and purchasing behavior prediction of high-valued customers
Authors:	唐子鈞;Tang, Tze-Chun
Contributors:	淡江大學統計學系碩士班陳景祥;Chen, Ching-Hsiang
Keywords:	RFM模型;叢型資料;連檢定;行為預測;RFM Model;Data clumps;Run test;Behavior Prediction
Date:	2016
Issue Date:	2017-08-24 23:43:50 (UTC+8)
Abstract:	大數據時代的來臨，已經顛覆以往用傳統人口變量(例如：性別、年齡、地區…等)做分析，取而代之的是每位顧客的過去各種紀錄。現今許多企業擁有龐大的顧客交易歷史資料，為了瞭解顧客的購買行為，最常被使用的是RFM模型，RFM模型可以簡單且有效的描述顧客的行為模式，並找出高價值的客戶，以利企業做出後續的決策。但是近年來越來越多學者提出不同指標修正RFM模型，例如將RFM模型修正為RFMC模型，新指標為叢(Clumpiness)探討每位顧客的購物模式，本研究以RFMC為出發點，提出另一個計算叢的方法並和過去學者的算法進行比較。最後再引用連(Run)的概念，預測顧客下一次的行為。 With the advent of big data era, instead of using demographic variables, such as gender, age, area, to analyze customers’ behavior, customers’ transaction history are used most frequently. Nowadays, many companies have huge amount of customers’ transaction data. In order to understand the behavior of customers, RFM (recency, frequency, and monetary) model is the most common method to use. RFM model is able to simply describe the mode of consuming behavior and find the high-valued customers so that the companies can make the right decisions. However, recent years more and more scholars proposed different method to modify the RFM model. For example, there is a new idea, called RFMC model, using “Clumpiness” to research the customers’ behavior. Therefore, the research in this paper used the RFMC model and a new idea to calculate the “Clumpiness”. Moreover, there is a comparison between this new idea and the other method in calculating “Clumpiness”. Finally, the concept of “Run” is cited to predict the next behavior of the customers.
Appears in Collections:	[Graduate Institute & Department of Statistics] Thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	161	View/Open

Loading...