臺灣中小企業製造業供應鏈分群與分類模型建立

Constructing a Clustering and Classification Model of Supply Chain Management for the Small and Medium Manufacturing Enterprises in Taiwan

蕭惠雯(Hui-Wen Hsiao)

指導教授 ： 顏憶茹

國立臺中科技大學/資訊與流通學院/流通管理系碩士班/碩士(2012年)

摘要

現今商業競爭環境激烈，企業不再是單獨的自主體系，而是整合上下游供應鏈，共同因應市場上的競爭。而供應鏈成員的績效可從多種角度進行深入探討，如策略面、流程面等，同時藉由強化供應鏈績效管理的知識與技術來提升企業競爭力。
本研究旨在以臺灣地區製造業廠商來建立供應鏈管理分群與分類模型並發展分類雛型表單。首先，透過因素分析來進行變數的縮減，再經由第一階段的華德法及第二階段的k-mean進行群集分析，將廠商分為三群。第一群:「進口型」廠商原物料進口比重較高，顯示此群對國外原物料及零組件具有依賴性，但對於原物料的掌控能力受到影響;第二群-「自主型」廠商則零進口，對於臺灣內部掌握度相較其他兩群優越;「穩健發展型」廠商部份原物料依賴進口，其庫存誤差率較差。
接著藉由C5.0建立規則，發展出分類雛型表單，可讓廠商以七題題項快速得知自身於供應鏈的定位，並依自身資源選取策略，以改善供應鏈效能，使得表單在實務運用上具更可行性。

關鍵字

C5 ； 供應鏈管理 ； 分類與分群

並列摘要

Currently the enterprise is no more an isolated system. It has to integrate the downstream and upstream supply chain to meet the competitive market. The performance of a supply chain member can be measured from various aspects, such as the strategic aspect and the processing aspect. It is critical for an enterprise to strengthen its knowledge and technology regarding the management of supply chain performance.
This study builds the clustering and classifying supply chain management model and develops the cluster prototype form based on the manufacturers located in Taiwan. First, this study uses factor analysis to delete the minor factor loading variables. Then by conducting the cluster analysis by the Ward’s and K-mean, this study classifies the manufacturers into three groups. Every group was named and the strategic recommendations were provided according to the group’s characteristics. The first group was named the ‘’Import Type’’ as it imports most of the raw materials and is much dependent on foreign suppliers. The second group has zero importation, and has better control of the raw material from Taiwan than the other two groups. This group was thus named the ‘’ Independent Type’’. The third group relies partly on imported raw material, and has poor inventory error rate. It was therefore named the ‘’Stable Development Type’’.
Then, this study generalizes the 47 items into eight rules by C5.0 and develops the classification prototype form. The classification prototype form can help the manufacturers quickly learn their own position in the supply chain, and thus the manufacturers can establish strategy based on their own resources to improve supply chain efficiency.

並列關鍵字

C5 ； supply chain ； clustering and classifying

參考文獻

邱秋婷(民96)。屬性導向方法應用於證券交易相對關係規則之挖掘(碩士論文)。國立中央大學，桃園縣。

陳叡宜(2001)。資訊分享對供應鏈績效影響之研究-從供應商角度探討(碩士論文)。元智大學，桃園縣。

董和昇(譯)(2010)。管理資訊系統概論(原作者:O’Brien, James A., Marakas, George M.)。臺北市:麥格羅希爾。(原著出版年:2010)

戴文禮(民97)。供應鏈績效評估與最適標竿之選擇(博士論文)。國立成功大學，臺南市。

Barak, A., & Gelbard, R. (2011). Classification by clustering decision tree-like classifier based on adjusted clusters. Expert Systems with Applications, 38(7), 8220-8228.