| Summary: | 碩士 === 輔仁大學 === 應用統計研究所 === 85 === With social and economic change, the market demand of credit card is on increase in Taiwan. Therefore, many banks are entering the credit card market. To cope with competition in the market, how will the issuing banks compete with each other? If the issuing banks make good marketing program they can to gain sustainable competitive advantage.
「Forecast」is the important tool to set up strategic plan, including market potential, sales forecast and so on. Therefore , the total expenditures of cardholders from 1994 to 1996 is used as a dependent variable to establish the sales forecast model in the research. Based on the data in Statistics Offices, Bureau of Monetary Affairs, Ministry of Finance, only VISA, AE, MASTER, DINERS and NCC will be discussed.
Two methods are used in the paper : one is regression analysis , the other is fuzzy regression analysis. Regression analysis is a statistical tool widely used to model the relationship among variables for the purpose of explanation or prediction, e.g., so that a dependent variable can be predicted from independent variables. This classical regression technique is useful in a non-fuzzy environment where the relationship among variable is crisply defined.
Fuzzy regression is a nonparametric method in the sense that the deviations between the observed valuse and the estimated values are assumed to dependent on indefiniteness/vagueness of the parameters which govern the system structure, not on its measurement errors. Because the actual data of the cardholders can''t be obtained, the fuzzy regression analysis will be applied to the research.
The consumption tendency of cardholders will be investigated in the research. Hopefully, the sales forecast model in this paper will help the issuing banks realize the factors affecting the cardholders'' consumption behaviors in order to make promotion decision. Besides, the result of sales forecast can offer the ideas of the sales quota and sales budget for the issuing banks.
|
|---|
