在半導體製造中，晶圓拋光過程產生破片是一個非常嚴重的問題，不僅是晶圓的成本相當昂貴，也會因為需處理機台問題和重新設置失去很多時間，更不用說是影響對顧客交貨的信譽。拋光機台相關的參數很多，但真正和產生晶圓破片有關的往往是集中在幾個少數關鍵因子上。本研究的目在於建立一個模型以辨識機台發生前之關鍵參數變化，以判斷機台可能破片。此辨識模式可為將來機台及時監控之用，使機台在可能破片之前及時停機以做處理，可以大幅降低成本損失，提升機台製程良率。
本研究以馬氏田口和邏輯迴歸方法，建立拋光機台發生破片的關鍵影響因子，並且進行績效指標以敏感度、明確度、總正確性、以及ROC (Receiver Operating Characteristic)分析曲面下的面積來比較此兩法的優劣，結果顯示馬氏田口判斷力較邏輯迴歸為佳。利用馬氏田口建立一套判斷拋光製程狀態的破片判斷機制。

In semiconductor manufacturing, wafer breakage during polishing processes is a very serious problem. Not only the cost of wafers are very expensive, much valuable time are also lost due to problem handling and machine re-setups – not to mention delivery credibility with customers Though there are many parameters associated with the wafer polisher, it is expected only a smaller number of parameters has some relationship with wafer breakage. The goal of this research is to identify a small number of relevant parameters and/or build some index to monitoring wafer breakage potential before it actually happens so that the problem can be handled to reduce wafer breakages.
This research used Mahalonobis-Taguchi System (MTS) and Logistic Regression (LR) methods to identify the key factors which are related to wafer breakage of Polisher Machine. Sensitivity, specificity, Accuracy and Area Under Curve(AUC) are used to as performance indices of the two methods. The result showed that MTS performed better than LR. The MTS indice was then chosen as a mechanism for monitoring wafer breakage potential of Polisher Machine.

參考資料
Berry, M. and Linoff, G.(1997). “Data mining techniques for marketing, sales and customer support”, John Wiley, New York.
Bahman P.T.and Willam H.H.(2002). ”A Multivariate Logistic Regression Equation to Screen for Diabetes, ”Diabetes Care, Vol. 25, No.11, pp.1999-2003.
Fayyad, U., Piatetsky-Shapiro, G. and Smyth, P. (1996). “From data mining to knowledge discovery: An overview.” Advances in Knowledge Discovery and Data Mining, pp. 1-35. AAAI/MIT Press.
Fayyad, U., Piatetsky-Shapiro, G. and Smyth, P. (1996). “The KDD Process for Extracting Useful Knowledge from Volumes of Data”, Communications of the ACM, Vol. 39, pp. 27-34.
Han, J. and Kamber, M. (2003). “Data mining - concepts and techniques”, Academic Press.
Hilden, J. (1991). “The area under the ROC curve and itscompetitors”, Med. Decis. Making, 11: 95-101.
Lusted, L. B. (1971). “Signal detectability and medical decision making”. Science 171: 1217-1219.
Preston,F. W. (1927).”The Theory and Design of Plate Glass Polishing Machine”, journal of the society of glass technology, Vol. 11, pp.214-247.
Rosner B (2005). Fundamentals of biostatistics. 6th ed. CA: Duxbury.
Su, C. T. and Hsiao, Y. H., (2007). “An Evaluation of the Robustness of MTS for Imbalanced Data.” IEEE Transactions on Knowledge and Data Engineering, Vol. 19, pp1321-1332.
Taguchi, G., Chowdhury, S., and Wu, Y., (2001). “The Mahanalobis-Taguchi System”.
Taguchi, G., Jugulum, R. (2002). “The Mahalanobis-Taguchi Strategy”, John Wiley & Sons, New York.
Verhulst FC, Koot HM (1992). “Child psychiatric epidemiology: concepts,
methods, and findings”. Newbury Park: SAGE Publications.
Woodall, W. H., Koudelik, R., Tsui, K. L., Kim, S. B., Stoumbos, Z. G. and Carvounis, C. P., (2003). “A Review and Analysis of the Mahalanobis-Taguchi System” Technometrics, Vol. 45, pp1-15.
Zarudi, I. and Han, B.S. (2003).”Deformation and Material Removal Rate in Polishing Silicon Wafers”, Journal of Materials Processing Technology, 140, pp.641-645.
李培瑞(2005)。半導體製程資料特徵萃取與資料挖礦。科技管理學刊，第七卷第一期，第137~160 頁。
吳鴻志(2002)。運用關聯法則分析異常WAT 良率之機台組合。碩士論文，中正大學會計學系。
周孟賢（2000）。化學機械研磨時控製程參數最佳化技術。碩士學位論文，國立中興大學機械工程學系。
莊達人（1999）。VLSI 製造技術。高立圖書有限公司，台北市。
張銘文(2006)。網版印刷錫膏膜厚預測類神經網路模型與規則庫之建立。碩士論文，中原大學機械工程學系。
曾韋霖(2003)。印刷電路板工作站內製程品質資料挖掘及建構異常診斷系統。碩士論文，元智大學工業工程與管理學系。
黃偉碩(2005)。利用貝氏分類與因子分析法於半導體製程錯誤偵測與診斷。碩士論文，中華大學科技管理學系
黃榮吉(2001)。應用資料探勘與本體論之半導體設備遠端診斷系統之開發。碩士論文，成功大學製造工程研究所碩博士班。
葉俊吾。運用類神經網路建構SMT 錫膏印刷製程品質管制系統。碩士論文，成功大學製造工程學系。
楊叢印(2003)。結合電過濾/電透析技術處理CMP廢水並同步產製電解水之研究。博士論文，中山大學。
羅正忠,張鼎張(譯)。Hong Xiao著(2002)。半導體製程技術導論。臺灣培生教育出版。
劉應興 (2003)。類別資料分析導論。華泰文化。
蕭宇翔(2005)。應用MTS於非平衡資料分析之穩健性研究-以行動電話檢測流程為例」。博士論文，交通大學工業工程與管理學系。
簡禎富、林鼎浩、徐紹鐘、彭誠湧(2001)。建構半導體晶圓允收測試資料挖礦架構及其實證研究。工業工程學刊，第十八卷第四期，第37~48 頁。
簡禎富等主編(2005)。半導體製造技術與管理。國立清華大學出版社，2005年5月初版。
蘇朝墩 (2006)。品質工程。中華民國品質學會，2006年9月四版。