隨著消費性電子產品的生命週期縮短，高科技產業因應產品變化快速而導致生產時間緊縮的壓力提高，這個壓力於印刷電路板產業尤其明顯，因為印刷電路板在半導體產業上扮演測試的角色，產品需更早一步被設計出來。印刷電路版為少量多樣的產品特性，其產品複雜程度亦高，如何監控各站別生產時間與估計總生產時間是我們所關注的問題。本研究提出資料挖礦架構，整合貝氏網路模型篩選出影響總週期時間之關鍵站別與並協助管理者列出間接影響因子，除此之外，亦提出時間推移模型，協助管理者計算在不同站別等候片數的基準下每片印刷電路板的週期時間。本研究透過與國內具指標性印刷電路板公司合作進行實證研究，從大量的資料中使用貝式網路篩選出關鍵影響站別並驗證結果；且依循時間推移模型方法，推估不同情境下的週期時間。貝氏網路模型結果得以協助工程師找尋影響週期時間的關鍵站別並瞭解間接影響因子；時間推移模型不僅能協助線上工程師瞭解每片電路板於各站別的生產時間，亦能依據不同的生產情境推估總週期時間，再者交於業務單位承諾顧客達交時間。

Competition in high tech industry forces the field to consider the ways to monitor the duration of cycle time and to keep produce efficiency within a budget. Particularly, Printed Circuit Board (PCB) industry is sensitive to this issue since their product characteristic is about small-volume and large-variety production. The product complexity of PCB is high, and its manufacturing processes of PCB go through thirty-six processes so how to monitor each station and to estimate the total cycle time are the issues we concerned. In this paper, we use data mining framework to build up a model for factors extraction and proposes a cycle time evolution table for estimation the cycle time. The Bayesian network extracts the main factors that significant influence on total cycle time and the cycle time evolution table estimate the total cycle time per piece of the board. This study cooperates with PCB company in Taiwan for empirical research. Proposed framework extracts critical stations which influence the total cycle time from huge data to validate the results. Furthermore, the engineers follow the results to find the indirect impact factor. On the other hand, the study also uses the cycle time evolution table on estimating cycle time. The results give decision makers a criterion on estimating cycle time and committing delivery day.

Table of Contents
Abstract ........................................................................................................................ iv
List of Tables ................................................................................................................ ii
List of Figures ..............................................................................................................iii
List of Notations ..........................................................................................................iii
Chapter 1 Introduction ................................................................................................ 1
1.1 Background and Motivation ........................................................................................ 1
1.2 Research Objective........................................................................................................ 2
1.3 Organization of Thesis .................................................................................................. 2
Chapter 2 Literature Review ...................................................................................... 3
2.1 PCB industry Introduction of Industry and domain knowledge .............................. 3
2.2 Data mining in PCB industry ....................................................................................... 5
2.3 Bayesian Network.......................................................................................................... 8
2.3.1 Bayes’ theorem ........................................................................................................ 8
2.3.2 The advantages and applications of Bayesian network ......................................... 10
Chapter 3 Research framework ............................................................................... 12
3.1 Problem definition ....................................................................................................... 13
3.2 Data preparation ......................................................................................................... 13
3.2.1 Data collection and preliminary scanning .............................................................. 14
3.2.2 Data cleaning, conversion and calculation ............................................................. 14
3.2.3 Product categories filter ......................................................................................... 15
3.2.4 Data segmentation.................................................................................................. 15
3.3 Model construction...................................................................................................... 15
3.3.1 Construction of Bayesian network and model validation ...................................... 16
3.3.2 Construction of cycle time evolution table ............................................................ 18
3.4 Results interpretation ................................................................................................. 21
Chapter 4 Case Study for PCB cycle time estimation ............................................ 23
4.1 Problem definition ....................................................................................................... 23
4.2 Data preparation ......................................................................................................... 24
4.2.1 Data collection and preliminary scanning .............................................................. 24
4.2.2 Data cleaning, conversion and calculation ............................................................. 26
4.2.3 Product categories filter ......................................................................................... 26
4.2.4 Data segamentation ................................................................................................ 26
4.3 Model construction...................................................................................................... 27
4.3.1 Construction of Bayesian network and model validation ...................................... 27
4.3.2 Construction of cycle time evolution table ............................................................ 31
4.4 Results interpretation ................................................................................................. 35
Chapter 5 Conclusion and future direction ............................................................. 38
5.1 Conclusion .................................................................................................................... 38
5.2 Future direction ........................................................................................................... 38
Reference .................................................................................................................... 39

Belbachir, A. N., Lera, M., Fanni, A., and Montisci, A. (2005), "An automatic optical inspection system for the diagnosis of printed circuits based on neural networks," Proceedings of Fourtieth IAS Annual Meeting. Conference Record of the 2005 Industry Applications Conference, 2005., 2-6 Oct. 2005.
Chang, P.-C., Liu, C.-H., and Fan, C.-Y. (2009), "Data clustering and fuzzy neural network for sales forecasting: A case study in printed circuit board industry," Knowledge-Based Systems, Vol. 22, No. 5, pp. 344-355.
Chang, P.-C., Wang, Y.-W., and Tsai, C.-Y. (2005), "Evolving neural network for printed circuit board sales forecasting," Expert Systems with Applications, Vol. 29, No. 1, pp. 83-92.
Chien, C.-F., Chen, S.-L., and Lin, Y.-S. (2002), "Using Bayesian network for fault location on distribution feeder," IEEE Transactions on Power Delivery, Vol. 17, No. 3, pp. 785-793.
Chien, C.-F., Liu, C.-W., and Chuang, S.-C. (2015), "Analysing semiconductor manufacturing big data for root cause detection of excursion for yield enhancement," International Journal of Production Research, pp. 1-13.
Chien, C. F. and Hsu, C. Y. (2014), "Data Mining for Optimizing IC Feature Designs to Enhance Overall Wafer Effectiveness," IEEE Transactions on Semiconductor Manufacturing, Vol. 27, No. 1, pp. 71-82.
Cooper, L. G. (2000), "Strategic marketing planning for radically new products," Journal of marketing, Vol. 64, No. 1, pp. 1-16.
Cui, G., Wong, M. L., and Lui, H.-K. (2006), "Machine learning for direct marketing response models: Bayesian networks with evolutionary programming," Management Science, Vol. 52, No. 4, pp. 597-612.
Friedman, N., Geiger, D., and Goldszmidt, M. (1997), "Bayesian network classifiers," Machine learning, Vol. 29, No. 2-3, pp. 131-163.
Haberle, K. R., Burke, R. J., and Graves, R. J. (2010), "Cycle time estimation models for printed circuit board design," International Journal of Production Research, Vol. 40, No. 4, pp. 1017-1028.
Haberle, K. R. and Graves, R. J. (2001), "Cycle time estimation for printed circuit board assemblies," IEEE Transactions on Electronics Packaging Manufacturing, Vol. 24, No. 3, pp. 188-194.
Kodek *, D. M. and Krisper, M. (2004), "Optimal algorithm for minimizing production cycle time of a printed circuit board assembly line," International Journal of Production Research, Vol. 42, No. 23, pp. 5031-5048.
Kuk Won, K., Young Jun, R., Hyung Suck, C., and Hyung Cheol, K. (2000), "A neural network approach to the inspection of ball grid array solder joints on printed circuit boards," Proceedings of Neural Networks, 2000. IJCNN 2000, Proceedings of the IEEE-INNS-ENNS International Joint Conference on, 2000.
Kwon, Y., Omitaomu, O. A., and Wang, G.-N. (2008), "Data mining approaches for modeling complex electronic circuit design activities," Computers & Industrial Engineering, Vol. 54, No. 2, pp. 229-241.
Laakso, T., Johnsson, M., Johtela, T., Smed, J., and Nevalainen, O. (2001). Estimating the Production Times in PCB Assembly, Turku Centre for Computer Science.
Ng, M. K. (2000), "A note on constrained k-means algorithms," Pattern Recognition, Vol. 33, No. 3, pp. 515-519.
Shaw, M. J., Subramaniam, C., Tan, G. W., and Welge, M. E. (2001), "Knowledge management and data mining for marketing," Decision Support Systems, Vol. 31, No. 1, pp. 127-137.
Sim, H., Choi, D., and Kim, C. O. (2014), "A data mining approach to the causal analysis of product faults in multi-stage PCB manufacturing," International Journal of Precision Engineering and Manufacturing, Vol. 15, No. 8, pp. 1563-1573.
Tien, F. C., Yeh, C. H., and Hsieh, K. H. (2004), "Automated visual inspection for microdrills in printed circuit board production," International Journal of Production Research, Vol. 42, No. 12, pp. 2477-2495.
Tseng, T.-L. B., Jothishankar, M. C., and Wu, T. T. (2004), "Quality control problem in printed circuit board manufacturing—An extended rough set theory approach," Journal of Manufacturing Systems, Vol. 23, No. 1, pp. 56-72.
Vainio, F., Maier, M., Knuutila, T., Alhoniemi, E., Johnsson, M., and Nevalainen, O. S. (2009), "Estimating printed circuit board assembly times using neural networks," International Journal of Production Research, Vol. 48, No. 8, pp. 2201-2218.
Vainio, F., Maier, M., Knuutila, T., Alhoniemi, E., Johnsson, M., and Nevalainen, O. S. (2010), "Estimating printed circuit board assembly times using neural networks," International Journal of Production Research, Vol. 48, No. 8, pp. 2201-2218.
Zhang, F. and Luk, T. (2007), "A Data Mining Algorithm for Monitoring PCB Assembly Quality," IEEE Transactions on Electronics Packaging Manufacturing, Vol. 30, No. 4, pp. 299-305.