應用聚合式階層分群法對產品家族分類之研究–以汽車零配件生產商之產品為例

魏良羽(Liang-Yu Wei)

指導教授 ： 曾富祥

國立中央大學/管理學院/工業管理研究所碩士在職專班/碩士(2014年)

摘要

製造業競爭日趨白熱化，大部份的客戶都希望降低庫存，減少資金積壓，採購的習慣由大批量採購，改變成少量多樣的採購。但客戶這樣的採購習慣對於生產單位帶來極大的不便，因此如何在客戶下單的習慣是少量多樣，和生產與採購零件的最低訂購量要求，雙方之間如何取得對生產廠商最佳的平衡，製造業如何在這樣的條件下使生產成本不至於過高成為一重要課題。

本研究針對一汽車零配件製造商廠商為個案公司，再藉由文獻探討與個案研究作為邏輯推導，利用聚合式階層分群法建構模型(Agglomerative Hierarchical Clustering)對產品的物料清單(Bill of Material, BOM)做分群，並透過個案公司產品資料，套入研究，以期將眾多產品有效分成數個家族，在日後面臨少量多樣的訂單時，在生產規劃、零件採購談判、降低生產成本上能有幫助。並也能幫助其他有類似困擾之公司解決其困擾。

關鍵字

聚合式階層分群法 ； 材料清單 ； 少量多樣

並列摘要

Businesses are getting more and more competitive, causing the majority of the companies today strive to eliminate the surplus cost of inventory in order to boost their cash flow. Therefore, companies purchase lower quantities of a greater variety of products to increase product diversity. This trend is a benefit to the buyers/customers, but it also creates the opposite effect to the manufacturers. To achieve the balance between customers’ buying criteria and suppliers’ minimum quantity requirements and how to keep the cost of production as unexpansive as possible has become an issue.

This research selects an automobile parts manufacturer to be the case company, and is supported by the literature review and case study to reach the logical conclusion. In addition, it adopts Agglomerative Hierarchical Clustering to collect numerous products of the case company from their bill of material, and place them into different subgroup based on the analysis of their historical data. The research is to develop the solution for case company to manage their future production planning to meet customers’ needs and establish their purchase negotiation strategy with the suppliers; meanwhile, the study can hopefully be valuable for other businesses that are encountering similar situation.

並列關鍵字

Agglomerative Hierarchical Clustering ； BOM

參考文獻

[2] A. K. Jain, M. N. Murty, P. J. Flynn. (1999). Data clustering: a review. ACM Computing Surveys (CSUR), Page 264-323 .

[7] Peter Henry Andrews Sneath, Robert R. Sokal. (1973). Numerical Taxonomy: The Principles and Practice of Numerical Classification. London, UK: W. H. Freeman.

[8] Karen L. Oehler, Robert M. Gray. (1995). Combining Image Compression and Classification Using Vector Quantization. IEEE Trans. Pattern Anal., Page 461-473.

[9] Michalski, Ryszard S. and Stepp, Robert E. (1983). Automated Construction of Classifications: Conceptual Clustering Versus Numerical Taxonomy. Pattern Analysis and Machine Intelligence, IEEE Transactions on, Page 396 - 410.

[10] Peter Henry Andrews Sneath, Robert R. Sokal. (1973). Numerical Taxonomy: The Principles and Practice of Numerical Classification. London, UK: W. H. Freeman.