臺灣博碩士論文加值系統

McFadden(1978) 提出GEV 模式做為解決一般之random utility 模式之分配模型, McFadden(1978) 同時提出NL 模式, 以建立層級結構的方式產生相關性, 解決了選項間的相關性問題, 但仍然有許多相關性問題尚未解決。McFadden(1978) 中之所以不採用更一般化的GEV 模式之原因是求MLE 解的困難。Small(1987) 對OGEV 模式採用兩階段MNL 求MLE 解之近似解, 但結果並不太理想。本文中建立了一個一般性的GEV 模式, 稱之為Total Weighted Logit 模式, 我們推導了模式中之數學性質, 並同樣採用Small(1987) 中使用之兩階段MNL 解求近似解。但由於其中權重參數之估計值尚未完成, 故結果仍不完善。

This paper is a new method to solve the problem of coffee purchased.

1 導論
2 Random Utility Model
2.1 Scanner Panel Data
2.2 符號定義與說明
2.3 咖啡購買資料說明
2.4 相關性來源探討
2.4.1 Carry-over effect
2.4.2 Random effect model
2.5 用MNL 模式分析咖啡購買資料
3 GEV 模式族
3.1 GEV 模式
3.2 GEV 模式族
3.3 我們的模式:Total Weighted Logit 模式
3.4 求我們的模式之MLE 解
4 用我們的模式處理咖啡的銷售數據
4.1 數據分析
5 討論
參考文獻

Abramson, C.,Andrews, R.L., Currim, I.S., and Jones M. (2000). ”Parameter Bias from Unobserved Effects in the Multinomial Logit Model of Consumer Choice ,” Journal of Marketing Research,37, 410-426.
Fish, K.E., Johnson, J.D., Dorsey R.E., and Blodgett, J.G. (2004). ”Using an Artificial Neural Network Trained with a Genetic Algorithm to Model Brand Share,” Journal of Business Research,57, 79-85.
Gelhausen,M.C. (2007). ”A Generalized Neural LogitModel for Airport and Access Model Choice in Germany,” Munich Personal RePEc Archive,4313,1-42.
Guadagni,P.M.,and Little,J.D.C. (1983). ”A Logit Model of Brand Choice Calibrated on Scanner Data,” Marketing Science ,2, 203-238.
Kalwani,M.U.,Yim, C.K.,Rinne,H.j.,and Sugita Y. (1990).” A Price Expectations Model of Customer Brand Choice,” Journal of Marketing Research,27, 251-262.
Mcfadden,D. (1978).” Modeling the Choice of Residential Location,” Spatial Interaction Theory and Planning Models,eds.A.Karlqvist,L.Lundquist,F.Snickers and J.Weibull,North-Holland, 75-96.
Small,Kenneth A. (1987).” A Discrete Choice Model for Ordered Alternatives,” Econometrica,55, 409-424.