臺灣博碩士論文加值系統

在半導體產業中，產品之良率高低將影響公司之營運成本與競爭力，故提升良率是每一間公司的重要目標，然而技術的進步固然重要，確保其產品之良率維持在應有之水準更為重要。由於半導體產業製程相當複雜，產品中之檢驗站往往是需經過數百個製程站才可執行，若其良率發生變異，要從中找出有問題的製程站，對於工程師而言為一大挑戰。
在現有文獻中，對於解決檢驗良率是否發生變異並找出其正確位置，尚未有一較佳的方式，因此開發偵測良率之自動化系統極為重要。在本篇論文中，將利用數學方式建立模型，提供偵測良率之方式，使工程師更有效率的解決有問題的製程站。
本篇論文所提供之方式，主要想法是來自於CART (Classification And Regression Tree)中之迴歸樹作法，並對其做一改良，進而將此應用至半導體產業界上。由於半導體產業中製造過程，常會出現離群值，其對於數學上建立模型為一困擾，因此在本文中對於離群值之出現，亦提供一方式來解決離群值對所建立之模型影響。
而在本文中，將會與2000年Wayne A. Taylor博士所發表的方法作比較。利用模擬的方式，建立均值平移之模型，模擬產品良率變動之情形，並且以偵測出其變異所發生之位置來比較其正確率。

In the semiconductor industry, a yield rate will affect the cost of business and the competitive power the company. Therefore, promoting a yield rate contributes to each company's profitable target. However, not only is the technical progress undoubtedly important, but a company’s guarantee that its product will have a standard yield rate is also important. Because the semiconductor industry’s system regulation is quite complex, a product must pass through hundreds of process stations to completely manufacture the product. After completing a system of ownership regulation, the product will be able to detect its yield rate. Therefore when the yield rate varies, it is an enormous challenge for engineers.
In current literature, there is no good way to solve the process of detecting whether to have variation and to discover a correct position. Therefore it is very important to develop an automatic system to detect the variation of the yield rate. In this paper, we will establish a model using mathematics, provide a way to detect the yield rate, and provide engineers a more effective solution to find the problem station.
The main ideal of this paper comes from CART (Classification And Regression Tree). This paper improves on it and then applies this method to the semiconductor industry. In the manufacturing process in the semiconductor industry, the regular session presents the outlier, and it is confusing to use mathematics in the model. Therefore the appearance of an outlier also provides a way to solve it.
Also, in this paper, our method will compare the accuracy rate with CPD (statistical Change-Point Detection), which was proposed by Dr. Wayne A. Taylor (2000a). Using a simulation, we set up models of the mean shift to simulate situations of changing product yield rates and use the detection of its varying position to compare its accuracy.

CHINESE ABSTRACT . I
ENGLISH ABSTRACT II
誌 謝 III
LIST OF FIGURES V
LIST OF TABLES . VI
CHAPTER 1: INTRODUCTION .1
1. 1 MOTIVATION AND OBJECTIVES 1
1. 2 THE PROCEDURE OF RESEARCH .4
1. 3 ORGANIZATION 5
CHAPTER 2: LITERATURE REVIEW 6
2. 1 USING CPD TO DETECT THE MEAN-SHIFT PROBLEM 6
2. 2 USING REGRESSION TREES TO DETECT THE MEAN-SHIFT PROBLEM .8
2. 2. 1 Introduction of regression trees 8
2. 2. 2 Partition 9
2. 2. 3 Pruning .10
2. 2. 4 The challenge of using regression trees to detect mean-shift 11
2. 2. 5 Cross-Validation 13
2.3 OUTLIER .15
CHAPTER 3: NEW METHOD .17
3. 1 INTRODUCTION OF A NEW METHOD .17
3. 2 THE INFLUENCE OF OUTLIERS 20
3. 3 FINDING DIFFERENT LEVEL MEAN SHIFTED BY MULTI-RESOLUTION .21
CHAPTER 4: EXPERIMENT 24
4. 1 NO SHIFT 25
4. 2 SHIFTED ONE TIME .26
4. 3 SHIFTED SEVERAL TIMES 28
4. 4 THE INFLUENCE OF DIFFERENT SCALES .30
CHAPTER 5: CONCLUSIONS 31
REFERENCES 33
APPENDIX 4.1 36
APPENDIX 4.2 38

[1] Abu-Taleb, A.A., Alawneh, A.J., and Smadi, M.M., Statistical analysis of recent changes in relative humidity in Jordan. American Journal of Environmental Sciences 3 (2), 2007, 75-77.

[2] Bergeret, F. and Le Gall, C., Yield Improvement using Statistical Analysis of Process Dates, IEEE Transactions on Semiconductor Manufacturing, Vol. 16, No. 3, 2003, 535-542.

[3] Besse P., Le Gall, C., Application and reliability of change-point analyses for detecting a defective in intragated circut manufacturing series, Communication in Statistics, Simulation and Computation , 2006.

[4] Breiman, L., Friedman, J.H., Olshen, R.A. and Stone, C.J., Classification and Regression Trees, Wadsworth, Belmont, California, 1984.

[5] Carslaw, D.C., Ropkins, K., and Bell, M.C., Change-Point Detection of Gaseous and Particulate Traffic-Related Pollutants at a Roadside Location, Environmental Science and Technology, Vol. 40. Issue 22, 2006, 6912-6918.

[6] Esposito, F., Malerba, D., and Semeraro, G., A Comparative Analysis of Methods for Pruning Decision Trees, IEEE Transactions on Pattern Analysis and Machine Intelligence, VOL. 19, NO. 5, 1997, 476-491

[7] Esposito, F., Malerba, D. and Semeraro, G., A Comparative Analysis of Methods for Pruning Decision Trees, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 19, No. 5, May 1997, 476-491.

[8] Kucera, J., Barbosa, P., Strobl, P., Cumulative sum charts - A novel technique for processing daily time series of MODIS data for burnt area mapping in Portugal, IEEE Proceedings Multitemp2007, Leuven, Belgium.

[9] Lai, T.L. Sequential change point detection in quality control and dynamical systems, J. Royal Statistical Society Soc. Ser. B 57, 1995, 613-658.

[10] Lavielle, M., Optimal segmentation of random processes, IEEE Transactions on signal processing 46, May 1998, 1365-1373.

[11] Neretti, N., Remondini, D., Tatar, M., Sedivy, J.M., Pierini, M., Mazzatti, D., Powell, J., Franceschi, C., and Castellani, G.C., Correlation analysis reveals the emergence of coherence in the gene expression dynamics following system perturbation, BMC Bioinformatics 2007, 8(Suppl 1):S16 (8 March 2007).

[12] Neter, J., Wasserman, W., and Kutner, M.H., Applied Linear Regression Models, Second Edition, Richard D. Irwin. Inc., Boston, Massachusetts, 1989.

[13] Smadi, M.M., Zghoul, A.A., A Sudden Change In Rainfall Characteristics In Amman, Jordan During The Mid 1950s, American Journal of Environmental Sciences 2 (3), 2006, 84-91.

[14] Taylor, W.A. Change-Point analysis: A powerful new tool for detecting changes, 2000. http://www.variation.com/cpa/tech/changepoint.html

[15] Williams, D., Kuhn, A., Kupsch, A., Tijssen, M., van Bruggen, G., Speelman, H.,
Hotton, G., Yarrow, K., and Brown, P., Behavioural cues are associated with modulations of synchronous oscillations in the human subthalamic nucleus. Brain, September 1, 2003; 126(9): 1975-1985.

[16] Windeatt T., Ardeshir G., An empirical comparison of pruning methods for ensemble classifiers, Proc. of Int. Conf Intelligent Data Analysis, Sept 13-15, 2001, Lisbon, Portugal, Lecture notes in computer science, Springer-Verlag, 208-217