臺灣博碩士論文加值系統

在現行管理作業上遇到品質異常逃脫時 ,多數人皆是頭痛醫頭 ,腳痛醫腳 .在20世紀迄今品管界持續推動著同步工程的概念 ,其目的是希望將客戶需求與品質同步設計在產品上 ,希冀品質能在設計端即先管控降低後續品質管理所產生的成本浪費 .
近年來類神經網路的分析手法方興未艾 ,該技術的發現是在資訊工程學系作資料挖礦時所誕生 ,該程式具有從龐大資料中找出規則性或關連性的強大能力 ,故本文研究案例將聚焦在傳統印刷電路板產業中的微鑽孔製程的製程能力指標 ,期望從其中找出符合的決策規則 ,過程則需藉由資料挖礦的建模手法找到高相關性的模型 ,並針對該模型主因子進行迴歸預測模型的建構 ,與運用羅吉斯模型的比較模式分析 ,以選擇出最低成本與品質不變的模型來建構新設計規則 .

Meets when the present management work the quality exceptionally escapes, the most people all are the headache medicine head, the foot painful medicine foot. The jurisdictional boundary was impelling the synchronized project concept continually until now in the 20th century, its goal is the hope the customer demand and the quality synchronization design on the product, hoped for that the quality can namely control first in the design end reduces the cost waste which the following quality administration center produces.
In recent years the kind of nerve network''s analysis technique was on the rise, this technology''s discovery was when the information project department made Data Mining was born, this program had from the huge material discovers the regularity or correlation formidable ability, therefore this article will study the case to focus in the traditional PCB industry Micro-drill system regulation process capability system, expected that from discovered the decision rule which tallied, the process must find the high relevant because of the Data Mining technique model, and aimed at this model main factor to use regression model to predict , with using the Logistic model''s comparison pattern analysis, chose the most low cost and the quality invariable model construction design rule.

中文提要 …………………………………………………………… i
英文提要 …………………………………………………………… ii
誌謝 …………………………………………………………… iii
目錄 …………………………………………………………… iv
圖目錄 …………………………………………………………… v
表目錄 …………………………………………………………… vii
一、 緒 論…………………………………………………… 1
1-1 研究動機與目的………………………………………… 1
2-2 研究背景………………………………………………… 2
3-3 研究方法………………………………………………… 3
二、 基本理論探討…………………………………………… 5
2-1 統計方法之迴歸基礎理論探討………………………… 5
2-2 製程管理指標 (Ca,Cp,Cpk)理論……………………… 7
2-3 資料挖礦理論…………………………………………… 11
三、 資料挖礦模型…………………………………………… 15
3-1 複合式一般線性模型 (Generalized linear mixed models) 15
3-2 CHAID 決策數…………………………………………… 16
3-3 羅吉斯迴歸模型 (Logistic regression model)…… 18
四、 鑽孔模型建立–因子討論 …………………………… 20
4-1 鑽孔作業流程簡介 …………………………………… 20
4-2 鑽孔孔偏製程管理指標探討…………………………… 23
4-3 鑽孔影響因子分析……………………………………… 24
五、 資料分析 ……………………………………………… 26
5-1 鑽孔設計分類規則的新舊比較………………………… 26
5-2 GLM 與CHAID主因子迴歸預測模型的計算…………… 30
5-3 GLM 與CHAID修正前預測模型的結果比較分析……… 38
5-4 GLM 與CHAID修正後預測模型的結果比較分析……… 42
六、 結 論…………………………………………………… 47
6-1 最佳建模模型比較結果………………………………… 47
6-2 GLM 模型設計規則分類決策…………………………… 48
6-3 未來改善空間…………………………………………… 49
參考文獻 …………………………………………………………… 50
附件一 Drill Raw Data………………………………………… 52

﹝1﹞ 線性迴歸：Neter, John 等著，劉應興譯：應用線性迴歸模型(Applied Linear Regression Models)。1996 年3 版，華泰書局。
﹝2﹞Statistic for Business and Economics : David R. Anderson , Sweeney , Williams ,Chen , 2006 ,THOMSON
﹝3﹞OPERATIONS MANAGEMENT ,Ninth Edition.: William J.Stevenson Edited. 2007. MC Graw Hill. pp. 446-473.
﹝4﹞Juran, J. M., Juran’s Quality Control Handbook, McGraw-Hill, New York (1974).
﹝5﹞品質管理，第二版 :簡聰海等編著，2001，全威圖書有限公司，台北，pp. 231-246。
﹝6﹞Lai K, Chen, Smiley W, Cheng, F. A. Spiring, 1988, “A new measure of process capability : Cpm ”Journal of Quality Technology , 20(3),pp.41-52.
﹝7﹞Pearn WL, Kotz S., Johnson NL. 1992, “Distributional and Inferential Properties of Process Capability Indices” Journal of Quality Technology , 24(4).
﹝8﹞Frawley, W. J. & Piatetsky-Shapiro, G. & Matheus, C. J.Knowledge Discovery Databases: An overview, in Knowledge Discovery in
Database, Cambridge, MA: AAAI/MIT, 1991, pp. 1-27.
[9] Fayyad, U. M. Data Mining and Knowledge Discovery(1996)：” Making Sense out of Data. IEEE Expert ” , pp 20-25.
[10] Berry, M. J. A.. & Linoff, G. S. Data Mining Techniques, John Wiley &
Sons, Inc. 1997, pp.407.
[11] Berson, A., & Smith, S. J., & Thearling, K. Building Data Mining Applications for CRM, Mc Graw-Hill, 1999, pp. 6-31.
﹝12﹞ Lewis, P. M. & Bernstein, A. & Kifer M. (2001). “ Database and Transaction Processing-An Application-Oriented Approach “ , pp. 645-663
﹝13﹞Panangaden P. and Taylor K. (1992).“ Concurrent Common Knowledge: defining agreement for asynchronous systems.” Distributed Computing 6,
pp. 73-93.
﹝14﹞Gao,J,X B.M. Manson, and P.Kyratsis(2000),”Implementation of concurrent engineering in the suppliers to the automotive industry,” Journal of Materials Processing Technology, pp. 107, 201-208
﹝15﹞BEDRICK, E. J., CHRISTENSEN, R. and JOHNSON, W. (1996). “A new perspective on priors for generalized linear models.” J.Amer. Statist. Assoc. 91
pp. 1450–1460.
﹝16﹞MCCULLAGH, P. and NELDER, J. A. (1989).“ Generalized Linear Models, 2nd ed. Chapman and Hall, London.”pp. 210-234
﹝17﹞Kass, G. V., “An Exploratory Technique for Investigating Large Quantities of Categorical Data”, Applied Statistics, Vol. 29, 1980, pp. 119-127.
[18] APPLIED LINEAR REGRESSION MODEL : Kutner ,Nachtsheim and Neter Fourth Edited. 2004. MC Graw Hill. pp. 555-568.
﹝19﹞電路板機械加工技術 : 台灣電路板協會編輯 , 童家慶發行 , 林定皓審稿. 2004 年初版，漢臨文化事業有限公司。
﹝20﹞劉朝勝 ,2002 ,“ 微鑽針之孔位偏移量的有限元素分析與實驗探討“, 國立中正大學機械研究所碩士論文 .
﹝21﹞陳興文 ,2002 ,“ 電路板微鑽針鑽尖面幾何設計之高速鑽孔性能研究“, 國立中正大學機械系，碩士論文。