應用層級貝氏模型在顧客異質性暨產品異質性之解讀

Application of Hierarchical Bayesian Model to Analysis of Customer Heterogeneity and Product Heterogeneity

廖子毅(Chih-Yi Liaw)

指導教授 ： 邱志洲 ； 吳忠敏

國立臺北科技大學/管理學院/商業自動化與管理研究所/碩士(2005年)

https://doi.org/10.6841/NTUT.2005.00047

摘要

在過去的行銷研究上我們可以發現，學者通常會將同一群顧客視為具有同質性的個體來進行分析。但是隨著行銷環境的改變，顧客的偏好逐漸趨向多元化，再加上每位顧客具備不同特性的結果，造成傳統將同群顧客視為同質性的概念已經不能符合實務之狀況，因著如此，所謂的顧客異質性（Customer Heterogeneity）概念乃因應而生。間接的也促成了近年來企業逐漸將如何塑造企業品牌，作為其主要競爭力的行銷優勢。為了成功達到塑造品牌差異化的競爭優勢，企業常須先針對消費者選擇品牌的異質性偏好進行了解。
為了能確切的提供企業了解如何成功塑造差異化優質品牌的策略，在本研究中，嘗試以行銷學的理論與觀點佐證層級貝氏模型的延展性與實用性。本論文的整體架構整理如下：第一章說明研究背景、動機以及研究架構；第二章利用零售產業的交易資料來進行消費者選擇品牌偏好的異質性分析，並根據實證結果提供企業了解顧客的消費偏好，以進一步提供整體解決方案，來塑造差異化的優勢品牌。第三章則是針對製造業中之物件衰退資料進行產品的異質性分析，並建構可靠度函數以提昇產品品質，協助廠商製造出更好、更具可靠性的產品。第四章則是將針對整個研究進行總結並提出後續的研究方向與建議。

關鍵字

層級貝氏模式 ； 品牌選擇模式 ； 可靠度函數 ； MCMC

並列摘要

We were allowed to discover in the past marketing research that the scholar had usually regarded the identical group customers as homogeneous individuals to carry on the analysis. But along with the marketing environment change, the customer preference progressively tends to multiplication. In addition, each customer has different characteristics. All of these lead the traditional concept of regarding the identical group customers as homogeneous individuals to be unable to match the reality. In consequence, the so-called concept of customer heterogeneity arises. This concept also contributes indirectly to the idea that enterprises use corporation brand formation as a major competence advantage. To achieve the competence advantage of successful brand differentiation, enterprises usually have to understand the preferences of customers for brand heterogeneity choice in advance.
This paper tries to test and verify the extensibility and practicability from the viewpoint of marketing to accurately provide enterprises with the strategy of successful corporation brand formation.
The overall structure of this paper is organized as follows. In section 1, research background, research motive and research structure are introduced. In section 2, we use real transaction data in the retailing industry for analysis of the choice of customer brand preference heterogeneity. Then, according to the practical results, customer preference is provided to propose an alternative for the formation of differentiated brand advantage. In section 3, an analysis for product differentiation of device degradation data in the manufacturing industry is performed. Also, we attempt to construct reliability function to improve production quality, facilitating manufacturers to produce better and more reliable products. In section 4, conclusion remarks and suggestions are made. Moreover, follow-up research directions are pointed out.

並列關鍵字

Hierarchical Bayesian Model ； Brand Choice Model ； Time-to-failure Distribution ； Markov Chain Monte Carlo

參考文獻

[2] Akama M. (2002), “Bayesian analysis for the results of fatigue test using full-scale models to obtain the accurate failure probabilities of the Shinkansen vehicle axle,” Reliability Engineering and System Safety, vol. 75, pp. 321-332.

[3] Allenby, Greg M. (1989), “A Unified Approach To Identify Estimating And Testing Demand Structures With Aggregate Scanner Data,” Marketing Science, vol. 8, no. 3. (Summer), pp. 265-280.

[4] Allenby, Greg M. and James L. Ginter (1995), “Using Extremes to Design Products and Segment Markets,” Journal of Marketing Research, vol. 32, pp. 392-403.

[5] Allenby, Greg M. and Peter J. Lenk (1994), “Modeling Household Purchase Behavior with Logistic Normal Regression,” Journal of American Statistical Association, vol. 89, pp. 1218-1231.

[6] Arora, N., G. M. Allenby, J. L. Ginter. (1998), “A hierarchical Bayesmodel of primary and secondary demand.” Marketing Scences, pp. 1729–1744.