臺灣博碩士論文加值系統

晶圓允收測試 (wafer acceptance test, WAT) 設置的目的在於對晶圓進行電性的量測，以做為晶圓出貨給代工客戶的依據。WAT 各項參數均訂有明確的規格，假如晶圓未能通過參數規格，因客戶對超過規格的晶圓有低良率的疑慮，所以通常的處置方式，就是報廢。如果可以事先預測 WAT 參數值，那麼即便發生 WAT 參數異常，工程師也可以盡快進行問題解決步驟，以減少 WAT 參數異常批的發生。

本研究之主要目的是發展一個以類神經網路為基礎之預測模型，用來預測半導體廠之 WAT 參數值。我們以某半導體業者之實際 WAT 參數及晶圓製造量測數據，來驗證預測模型之正確性。

此研究利用平均絕對誤差百分比 (mean absolute percent error, MAPE) 作為主要之評估指標，比較類神經網路與傳統迴歸分析之預測正確性。我們採用之類神經網路包含徑向基函數網路與倒傳遞網路。研究結果顯示，三種預測模型之 MAPE 均低於 2.5%，優於公司內現行之預測方法，其中以倒傳遞網路所建立的預測模型具有最佳的預測效果，徑向基函數網路則又優於傳統迴歸分析。

Wafer acceptance test (WAT) results are the basis of shipping wafers to foundry customers. If wafers fail in the criteria of WAT, the customer will reject them due to low yield. However, if we can predict the results of WAT during wafer process, engineers could conduct problem solving process sooner the WAT abnormality resulted. As a result, the quantity of WAT abnormal lots will be effectively reduced.

The main purpose of this study is to develop a neural network-based prediction model for WAT. Real WAT data and wafer process measurement data are employed to verify our proposed prediction model.

Two ANN models are considered in this research, namely RBFN and BPN. The traditional regression analysis is used as a benchmark for comparison. The mean absolute percent error (MAPE) is used as the primary performance measure in this research. A comparative study shows that the MAPE values of three prediction models are less than 2.5% and better than current method used in the company. BPN has the best overall performance, followed by RBFN and regression model.

中文摘要i
英文摘要ii
誌謝iii
目錄iv
圖目錄vi
表目錄vii

第一章 緒論1
1.1研究背景與動機1
1.2研究目的2
1.3研究方法2
1.4研究架構3
1.5研究流程4

第二章 文獻探討6
2.1資料探勘6
2.2類神經網路8
2.3徑向基函數網路11
2.4倒傳遞網路14
2.5迴歸分析17
2.6半導體製程21
2.6.1 晶圓製造過程 21
2.6.2 封裝測試過程 25
2.7半導體製程資料架構 26
2.7.1 製造執行系統 27
2.7.2 工程資料庫 28

第三章 研究方法 32
3.1半導體資料探勘技術研究架構 32
3.2半導體廠現況 33
3.3資料探勘前置處理 34
3.3.1 問題定義 35
3.3.2 資料萃取 35
3.3.3 資料準備 36
3.3.4 資料處理 38
3.4建立資料探勘模型 40

第四章 實驗結果與分析 43
4.1實驗規劃 43
4.2徑向基函數網路實驗 44
4.3倒傳遞網路實驗 53
4.4迴歸分析實驗 58
4.5三種方法之預測能力比較 61
4.6本研究與公司預測方法比較 63

第五章 結論與建議 64
5.1研究結論 64
5.2後續研究與建議 65

參考文獻 66

一、英文部分

1.Berry, M. J. A. and Linoff, G., Data Mining Techniques: for Marketing, Sales and Customer Support, John-Wiley & Sons Inc., New York (1997).
2.Berson, A., Smith, S. and Thearling, K., Building Data Mining Application for CRM, McFraw-Hill Inc., New York (2000).
3.Bishop, C., “Neural networks for pattern recognition,” Oxford University Press, (1995).
4.Brachman, R. J., Khabaza, T., Kloesgen, W., Piatetsky-Shapiro, G. S. and Simoudis, E.,”Mining business databases,” Communication of the ACM, 39(11), 42-48 (1996).
5.Cha, I. and Kassam, S. A., ”Channel equalization using adaptive complex radial basis function network,” IEEE Journal on Selected Areas in Communications, 13, 122-131 (1995).
6.Cybenko, G., ”Approximation by superposition of a sigmoidal function,” Mathematics of Control, Signals, and Systems, 2, 303-314 (1989).
7.Davies, P. C., “Design issues in neural network development,” Neurovest Journal, 5, 21-25 (1994).
8.Eksioglu, M., Fernandez, J. E. and Twomey, J. M., “Prediction peak pinch strength: artificial neural networks vs. regression,” International Journal of Industrial Ergonomics, 18, 431-441 (1996).
9.Fahlman, S. E. and Lebiere, C., “The cascade-correlation learning architecture. In: D. Touretzsky (Ed.),” Advance in Neural Information Processing, 598-605 (1990).
10.Fayyad, U. M.and Piatetsky, G. S., “The KDD process for extracting useful knowledge from volumes of data,” Commun ACM, 39(11), 27-34 (1996).
11.Fish, K. E., Barnes, J. H. and Aiken, M. W., “Artificial neural networks: a new methodology for industrial market segmentation,” Industrial Marketing Management, 24, 431-438 (1995).
12.Freeman, M. L. and Skapura, D. M., Neural Networks Algorithms, Applications and Programming Techniques, Addision-Wesley, New York (1992).
13.Goh, B. H., “Residential construction demand forecasting using economic indicators: a comparative study of artificial neural networks and multiple regression,” Construction Management and Economics, 14, 25-34 (1996)
14.Golub, G. and Kahan, W., “Calculating the singular values and pseudo-inverse of a matrix,” SIAM Numerical Analysis, B 2(2), 205-224 (1965).
15.Ham, F. M. and Kostanic, I., Principle of Neurocomputing for Science and Engineering, McGraw-Hill Inc., New York (2000).
16.Hand, D., Mannila, H. and Smyth, P., “Principles of data mining,” MIT Press, Cambridge, MA (2001).
17.Haykin, S., Neural Networks-A Comprehensive Foundation, Prentice-Hall, New York, (1994).
18.Hertz, J., Krogh, A. and Palmer, R. G., Introduction to the Theory of Neural Computation, Addision-Wesley, New York (1991).
19.Hornik, K., Stinchcombe, M. and White, H.,”Multilayer feedforward networks are universal approximators,” Neural Networks, 2, 359-366 (1989).
20.Hush, D. R., Salas, J. M. and and Horne, B., ”Error surfaces for multilayer perceptrons,” IEEE Transactions on Systems, Man and Cybernetics, 22 (1992).
21.Hush, D. R. and Horne, B., ”Progress in supervised neural networks: what’s new since Lippmann,” IEEE Signal Processing Magazine, January (1993).
22.Kim, K. B., Park, J. B., Choi, Y. H. and Chen, G., ”Control of chaotic dynamical systems using radial basis function network approximators,” Information Science, 130, 165-183 (2000).
23.Lapedes, A., Farber, R., “How neural networks. in: D. Z. Anderson (Ed.),” Neural and Information Processing System, American Institute of Physics, 442-556 (1988).
24.LeCun, Y., Denker, J. S. and Solla, S. A., “Optimal brain damage. in: D. Touretzsky (Ed.),” Advance in Neural Information Processing, 598-605 (1990).
25.Leonard, J. A. and Kramer, M. A., “Radial basis function networks for classifying process faults,” IEEE Control Systems Magazine, 11, 31-39 (1991).
26.Manoharran, S. C. ,Veezhinathan, M. and Ramakrishnan, S., ”Comparison of two ANN methods for classification of spirometer data,” Measurement Science Review, Vol.8, Sec.2, No.3 (2008)
27.McCulloh, W. S. and Pitts, W., “A logical calculus of the ideas immanent in nervous activity,” Bulletin of Mathematical Biophysics, 5, 115-133 (1943).
28.MESA International white Paper No.6 (1997).
29.Minsky, M. L. and Papert, S. A., “Perceptrons,” MIT Press Cambridge, MA (1969).
30.Montgomery, D. C., Design and Analysis of Experiment, 6th, John-Wieley, Chapter 10, 179-193 (2005).
31.Moody, J. and Darkin, C. J., “Fast learning in networks of locally-tuned processing units,” Neural Computation, No. 1(2), 281-294 (1989).
32.Park, J. and Sandberg, I. W., ”Universal approximation using radial basis function networks,” Neural Computation, 3(2), 246-257 (1991).
33.Rumehhart, D. E., Hinton, G. E. and Willaims R. J., “Parallel distributed processing,” MIT Press, Cambridge, MA, 1, 318-364 (1986).
34.Sahin, F., “A radial basis function approach to a color image classification problem in a real time industrial application,” Master''s Thesis, etd-6197-223641 (1997).
35.Sahoo, G. B., Ray, C., Wang, J. Z., Hubbs, S. A., Song, R., Jasperse, J. and Seymour, D., ”Use of artificial neural networks to evaluate the effectiveness of riverbank filtration,” Water Research, 39, 2505-2516 (2005).
36.Tanis, E. A. and Hogg, R., Hogg Probability and Statistical Inference, 7th ed., Prentice Hall , New York (2005).
37.Vellido, A., Lisboa, P. J. G. and Vaughan, J., “Neural networks in business: a survey of applications,” Expert Systems with Applications, 17, 51-70 (1999).

二、中文部分

1.莊達人，「VLSI製造技術」，高立圖書股份有限公司，台北，1990。
2.姜庭隆，「半導體製程」，滄海書局，台中，2001。
3.葉怡成，「神經網路模式應用與實作」，儒林圖書有限公司，台北 2004。
4.張詠棨，「半徑基底函數(RBF)類神經網路應用於LED晶圓缺陷檢測」，雲林科技大學資工所碩士論文，雲林，2006。
5.孔祥竹，「應用類神經網路減少TFT-LCD產品測試項目之研究」，交通大學工管所碩士論文，新竹，2007。

三、網路部分

1.半導體製造流程圖，http://www.necel.com/fab/en/flow.html
2.維基百科，類神經網路歷史，http://zh.wikipedia.org/wiki/.
3.MES Introduction, http://articles.e-works.net.cn/mes/Article35593.htm