頁數：52 頁 校所組別：國立中央大學管理學院工業管理研究所碩士在職專班 畢業時間及提要別：一百一十五學年度第二學期碩士論文提要 研究生：向家民 指導教授：王啓泰博士 近年隨著全球人工智慧與雲端數據中心的爆發性成長，網路設備是 AI 時代的 基礎建設。網通產品屬於系統整合型設備，生產效率是企業維持核心競爭力的關 鍵。但實際生產效率受到多項變數影響，容易忽略因素間的交互關係，導致難以聚 焦根本問題。因此，本研究目標是降低資料維度並找出影響效率的關鍵因子。 本研究運用主成分分析法，以全球網通領導廠商 D 集團之中壢生產工廠為研 究對象，蒐集整理其產線日常運作數據。研究架構依據製造管理之人、機、料、法、 環五大面向建構完整的產線資料指標體系。並透過 KMO 樣本適切性檢定，進一步 運用主成分分析法進行資料降維，藉此釐清各效率指標間的關聯，並萃取出最具 代表性的潛在因子群組 。 研究結果顯示，透過主成分分析法能成功找出關鍵主成分，並界定出對整體產 線效率影響程度最高的核心因素，並依據其解釋變異率確立了改善的優先順序。 此實證結果可在資源限制下，精準地識別產線關鍵因子。且除能有效優化個案公 司之產線運作結構與瓶頸管理外，亦可作為其他電子製造業未來導入數據化管理 與智慧製造決策支持的重要科學依據 。;Total page number: 52 Program and university: MBA， Graduate Institute of Industrial Management， National Central University Graduation semester and year: Spring 2026 Student: Chia-Ming Hsiang Thesis advisor: Dr. Chi-Tai Wang With the rapid growth of AI and cloud data centers， network communication equipment has become an essential infrastructure in the AI era. As network communication products are highly integrated systems， production efficiency plays a critical role in maintaining a company′s competitiveness. However， production efficiency is affected by multiple factors， and the relationships among these factors are often difficult to identify. Therefore， this study aims to reduce data dimensionality and identify the key factors affecting production line efficiency. This study applies Principal Component Analysis (PCA) to investigate the production line of manufacturing plant of Group D， a global leader in the network communication industry. Daily operational data were collected and analyzed. Based on the five dimensions of manufacturing management—Man， Machine， Material， Method， and Environment—a comprehensive production performance indicator system was established. After verifying data suitability using the Kaiser-Meyer-Olkin (KMO) measure， PCA was employed to reduce data dimensionality， identify relationships among efficiency indicators， and extract the most representative latent factor groups. The results show that PCA can effectively identify the key factors influencing production line efficiency and determine improvement priorities based on the explained variance of each principal component. The proposed approach helps manufacturers identify critical factors and improve production efficiency. In addition， the findings can serve as a useful reference for other electronics manufacturing companies seeking to implement datadriven management and smart manufacturing practices.
