簡易檢索 / 詳目顯示

研究生: 陳建智
Chen, Chien-Chih
論文名稱: 使用相依虛擬樣本從小樣本中學習更多資訊
Employing Dependent Virtual Samples for Learning More Information from Small Datasets
指導教授: 利德江
Li, Der-Chiang
學位類別: 博士
Doctor
系所名稱: 管理學院 - 工業與資訊管理學系
Department of Industrial and Information Management
論文出版年: 2011
畢業學年度: 100
語文別: 英文
論文頁數: 79
中文關鍵詞: 小樣本學習虛擬樣本資訊擴散相關係數
外文關鍵詞: small dataset, virtual sample, information diffusion, correlation coefficient
相關次數: 點閱:84下載:2
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 為因應市場多變的需求,工業產品的生命週期變得越來越短,尤其是電子產業。對於資本密集的電子產業而言,降低新產品試產次數以維持設備高稼動率是一個策略方向,然而如何從少量樣本中取得更多的製程資訊,是一個極具意義的研究課題。雖然虛擬樣本產生法藉由增加訓練樣本的方式為小樣本學習提供了一個有效的方法,然而如整體趨勢擴散技術(MTD)並未考量屬性間的相關性,而擴散神經網路僅能在高屬性相關性時方能運行。因此,本研究基於空間分割的概念,以迴歸分析為基礎,並結合模糊理論與可能性評估機制,發展一套考量原始資料屬性間相關性的虛擬樣本逐步產生法,以改善建模工具對小樣本的學習效能。文中先發展以探討兩變數間模糊相關性的二維模式,而後透過兩兩變數間模糊相關性的傳遞,進而衍化出多維模式。為具體呈現知識擷取能力的改善,本研究採用M5'模式樹做為建模工具,此外亦引用MTD進行倒傳遞類神經網路對小樣本的學習改善之比較。本論文以兩筆液晶面板產業的試產資料進行研究,結果發現所產出之虛擬樣本不僅可有程度地保有小樣本的模式行為,除可顯著地改善學習效能外,亦可擷取出更多資訊。同時在方法的比較上,整體而言亦較MTD為佳。

    Product life cycles are becoming shorter to meet the various demands of customers, especially in the electronic industry. For most firms in this capital intensive industry, it has thus become an important strategy to reduce pilot runs when new products are being phased-in in order to maintain a high equipment running rate, and to achieve this it is necessary to obtain more process information from a limited number of samples. While virtual sample generation (VSG) algorithms have shown their effectiveness in addressing this problem, the mega-trend-diffusion (MTD) technique does not consider the relations between attributes, and the diffusion neural network cannot work unless the linear correlation coefficient is higher than 0.9. Accordingly, this research develops a more effective VSG procedure based on the concepts of space partition, regression analysis, and fuzzy techniques to improve the learning performance of modeling tools. In this study, a two-dimensional model is developed by fuzzifying the correlation between two variables, and then a multi-dimensional model is further derived based on it by stepwise delivering the fuzzified correlations among attributes. In this research, the M5’ model tree is employed to concretely represent the information acquisition capability, while a back-propagation neural network (BPN) is also used to compare the proposed procedure with MTD. Two real datasets taken from a thin-film transistor liquid-crystal display (TFT-LCD) manufacturer are examined in the experiments. The results show that the samples created by the proposed procedure can retain a certain level of the original data behavior, and the information acquisition capability of M5’ and the learning perform of the two modeling tools can be significantly improved. In addition, the proposed procedure also outperforms MTD in certain diffusion levels in the two cases.

    摘要...............................................................................I ABSTRACT..................................................................II 誌謝.............................................................................III TABLE OF CONTENTS...............................................IV LIST OF FIGURES.......................................................VI LIST OF TABLES.......................................................VIII 1. INTRODUCTION..........................................................1  1.1 Backgrounds..............................................................1  1.2 Motivation.................................................................3  1.3 Objectives..................................................................4  1.4 Organization..............................................................5 2. LITERATURE REVIEW................................................6  2.1 Related Works...........................................................6    2.1.1 Virtual Sample Generation Algorithms…........6    2.1.2 Small Dataset Learning Approaches…...........15  2.2 The M5’ Model Tree.............................................17    2.2.1 Input Attribute Preprocessing.........................18    2.2.2 Data-partition Process....................................19    2.2.3 Regression Modeling.....................................19    2.2.4 Prediction Smoothing....................................20    2.2.5 The Implementation Procedure of M5'............20 3. METHODOLOGY.......................................................22  3.1 The Two-dimensional Model....................................22    3.1.1 Domain Estimation.......................................23    3.1.2 Value Filling................................................29    3.1.3 Sample Set Forming.....................................32    3.1.4 Summary of the Implementation Procedure....32  3.2 The Multi-dimensional Model..................................38    3.2.1 The Steering Variable Determination Process..40    3.2.2 The Initial Value Generation Process..............43    3.2.3 The Value Tuning Process.............................45 4. EXPERIMENTAL STUDY..........................................47  4.1 The Experimental Environment................................47  4.2 Manufacturing Cases in TFT-LCD Pilot Runs...........48    4.2.1 Case I: Shifting Forecasting...........................49    4.2.2 Case II: Photo-Spacer Height Forecasting.......62  4.3 Further Study and Discussions..................................68    4.3.1 The Knowledge Obtained..............................68    4.3.2 The Influence of Diffusion Level...................69 5. CONCLUSIONS.........................................................71  5.1 Conclusion.............................................................71  5.2 Future work............................................................72 REFERENCES................................................................73

    Anthony, M., Biggs, N. (1997). Computational Learning Theory. Cambridge University Press.
    Bhattacharya, B., & Solomatine, D. P. (2006). Machine learning in sedimentation modelling. Neural Networks, 19(2), 208-214.
    Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (1984). Classification and Regression Trees. Wadsworth and Brooks.
    Chan, K. Y., Kwong, C. K., & Tsim, Y. C. (2010). A genetic programming based fuzzy regression approach to modelling manufacturing processes. International Journal of Production Research, 48(7), 1967-1982.
    Chao, G. Y., Tsai, T. I., Lu, T. J., Hsu, H. C., Bao, B. Y., Wu, W. Y., et al. (2011). A new approach to prediction of radiotherapy of bladder cancer cells in small dataset analysis. Expert Systems with Applications, 38(7), 7963-7969.
    Dobra, A., & Gehrke, J. E. (2002). SECRET: A Scalable Linear Regression Tree Algorithm. Proc. Eighth ACM SIGKDD Int’l Conf. Knowledge Discovery and Data Mining, 481-487.
    Efron, B., & Tibshirani, R. J. (1993). An Introduction to the Bootstrap: New York: Chapmen & Hall.
    Frank, E., Wang, Y., Inglis, S., Holmes, G., & Witten, I. H. (1998). Technical note: Using model trees for classification. Machine Learning, 32(1), 63-76.
    Guo, G. D., & Dyer, C. R. (2005). Learning from examples in the small sample case: Face expression recognition. IEEE Transactions on Systems Man and Cybernetics Part B-Cybernetics, 35(3), 477-488.
    Hong, T. P., Tseng, L. H., & Chien, B. C. (2010). Mining from incomplete quantitative data by fuzzy rough sets. Expert Systems with Applications, 37(3), 2644-2653.
    Huang, C. F. (1997). Principle of information. Fuzzy Sets and Systems, 91(1), 69-90.
    Huang, C. F., & Moraga, C. (2004). A diffusion-neural-network for learning from small samples. International Journal of Approximate Reasoning, 35(2), 137-161.
    Huang, C. J., Wang, H. F., Chiu, H. J., Lan, T. H., Hu, T. M., & Loh, E. W. (2010). Prediction of the Period of Psychotic Episode in Individual Schizophrenics by Simulation-Data Construction Approach. Journal of Medical Systems, 34(5), 799-808.
    Ivănescu, V. C., Bertrand, J. W. M., Fransoo, J. C., & Kleijnen, J. P. C. (2006). Bootstrapping to solve the limited data problem in production control: an application in batch process industries. Journal of the Operational Research Society, 57(1), 2-9.
    Jang, J. S. R. (1993). ANFIS: adaptive-network-based fuzzy inference system. IEEE Transactions on Systems, Man and Cybernetics, 23(3), 665-685.
    Jennrich, R. I., & Schluchter, M. D. (1986). Unbalanced repeated-measures models with structured covariance matrices. Biometrics, 42(4), 805-820.
    Karalic, A. (1992). Employing linear regression in regression tree leaves. Proceedings of the 10th European Conference on Artificial Intelligence, 440-441.
    Kucuksille, E. U., Selbas, R., & Sencan, A. (2009). Data mining techniques for thermophysical properties of refrigerants. Energy Conversion and Management, 50(2), 399-412.
    Kuo, Y., Yang, T., Peters, B. A., & Chang, I. (2007). Simulation metamodel development using uniform design and neural networks for automated material handling systems in semiconductor wafer fabrication. Simulation Modelling Practice and Theory, 15(8), 1002-1015.
    Laird, N. M., & Ware, J. H. (1982). Random-effects models for longitudinal data. Biometrics, 38(4), 963-974.
    Lanouette, R., Thibault, J., & Valade, J. L. (1999). Process modeling with neural networks using small experimental datasets. Computers & Chemical Engineering, 23(9), 1167-1176.
    Li, D., Gu, H., & Zhang, L. Y. (2010a). A fuzzy c-means clustering algorithm based on nearest-neighbor intervals for incomplete data. Expert Systems with Applications, 37(10), 6942-6947.
    Li, D. C., Chang, F. M. M., & Chen, K. C. (2010b). Building reliability growth model using sequential experiments and the Bayesian theorem for small datasets. Expert Systems with Applications, 37(4), 3434-3443.
    Li, D. C., Chen, C. C., Chang, C. J., & Chen, W. C. (2011a). Employing Box-and-Whisker plots for learning more knowledge in TFT-LCD pilot runs. International Journal of Production Research (In Press). DOI: 10.1080/00207543.2011.555430.
    Li, D. C., Chen, C. C., Chang, C. J., & Lin, W. K. (2012). A Tree-based-Trend-Diffusion prediction procedure for small sample sets in the early stages of manufacturing systems. Expert Systems with Applications, 39(1), 1575-1581.
    Li, D. C., Chen, C. C., Chen, W. C., & Chang, C. J. (2011b). Employing dependent virtual samples to obtain more manufacturing information in pilot runs. International Journal of Production Research (In Press).
    Li, D. C., Chen, L. S., & Lin, Y. S. (2003). Using Functional Virtual Population as assistance to learn scheduling knowledge in dynamic manufacturing environments. International Journal of Production Research, 41(17), 4011-4024.
    Li, D. C., Fang, Y. H., Lai, Y. Y., & Hu, S. C. (2009a). Utilization of virtual samples to facilitate cancer identification for DNA microarray data in the early stages of an investigation. Information Sciences, 179(16), 2740-2753.
    Li, D. C., Hsu, H. C., Tsai, T. I., Lu, T. J., & Hu, S. C. (2007a). A new method to help diagnose cancers for small sample size. Expert Systems with Applications, 33(2), 420-424.
    Li, D. C., & Lin, Y. S. (2006). Using virtual sample generation to build up management knowledge in the early manufacturing stages. European Journal of Operational Research, 175(1), 413-434.
    Li, D. C., Lin, Y. S., & Huang, Y. C. (2009b). Constructing marketing decision support systems using data diffusion technology: A case study of gas station diversification. Expert Systems with Applications, 36(2), 2525-2533.
    Li, D. C., Liu, C. W., Fang, Y. H., & Chen, C. C. (2010c). A yield forecast model for pilot products using support vector regression and manufacturing experience-the case of large-size polariser. International Journal of Production Research, 48(18), 5481-5496.
    Li, D. C., Liu, C. W., & Hu, S. C. (2010d). A learning method for the class imbalance problem with medical data sets. Computers in Biology and Medicine, 40(5), 509-518.
    Li, D. C., Tsai, T. I., & Shi, S. (2009c). A prediction of the dielectric constant of multi-layer ceramic capacitors using the mega-trend-diffusion technique in powder pilot runs: case study. International Journal of Production Research, 47(1), 51-69.
    Li, D. C., Wu, C. S., & Chang, F. M. M. (2005). Using data-fuzzification technology in small data set learning to improve FMS scheduling accuracy. International Journal of Advanced Manufacturing Technology, 27(3-4), 321-328.
    Li, D. C., Wu, C. S., Tsai, T. I., & Chang, F. M. M. (2006). Using mega-fuzzification and data trend estimation in small data set learning for early FMS scheduling knowledge. Computers & Operations Research, 33(6), 1857-1869.
    Li, D. C., Wu, C. S., Tsai, T. I., & Lina, Y. S. (2007b). Using mega-trend-diffusion and artificial samples in small data set learning for early flexible manufacturing system scheduling knowledge. Computers & Operations Research, 34(4), 966-982.
    Li, D. C., & Yeh, C. W. (2008). A non-parametric learning algorithm for small manufacturing data sets. Expert Systems with Applications, 34(1), 391-398.
    Li, D. C., Yeh, C. W., & Li, Z. Y. (2008). A case study: The prediction of Taiwan's export of polyester fiber using small-data-set learning methods. Expert Systems with Applications, 34(3), 1983-1994.
    Li, R. M., Yao, J. Q., Shang, H. J., & Ruan, D. (2010e). An optimal model of information diffusion principles to risk and decision analysis of breast cancer morbidity. Soft Computing, 14(12), 1297-1303.
    Lin, Y. S., & Li, D. C. (2010). The Generalized-Trend-Diffusion modeling algorithm for small data sets in the early stages of manufacturing systems. European Journal of Operational Research, 207(1), 121-130.
    Liu, X. P., Zhang, J. Q., Cai, W. Y., & Tong, Z. J. (2010). Information diffusion-based spatio-temporal risk analysis of grassland fire disaster in northern China. Knowledge-Based Systems, 23(1), 53-60.
    Loh, W. Y. (2002). Regression trees with unbiased variable selection and interaction detection. Statistica Sinica, 12(2), 361-386.
    Niyogi, P., Girosi, F., & Poggio, T. (1998). Incorporating prior information in machine learning by creating virtual examples. Proceedings of the IEEE, 86(11), 2196-2209.
    Oniśko, A., Druzdzel, M. J., & Wasyluk, H. (2001). Learning Bayesian network parameters from small data sets: application of Noisy-OR gates. International Journal of Approximate Reasoning, 27(2), 165-182.
    Papari, M. M., Yousefi, F., Moghadasi, J., Karimi, H., & Campo, A. (2011). Modeling thermal conductivity augmentation of nanofluids using diffusion neural networks. International Journal of Thermal Sciences, 50(1), 44-52.
    Quinlan, J. R. (1992). Learning with Continuous Classes. Paper presented at the Proceedings Australian Joint Conference on Artificial Intelligence, World Scientific, Singapore.
    Thomas, M., Kanstein, A., & Goser, K. (1997). Rare fault detection by possibilistic reasoning. Computational Intelligence - Theory and Applications, 1226, 294-298.
    Tsai, T. I., & Li, D. C. (2008a). Approximate modeling for high order non.-linear functions using small sample sets. Expert Systems with Applications, 34(1), 564-569.
    Tsai, T. I., & Li, D. C. (2008b). Utilize bootstrap in small data set learning for pilot run modeling of manufacturing systems. Expert Systems with Applications, 35(3), 1293-1300.
    Vapnik, V. N. (2000). The Nature of Statistical Learning Theory: Springer, New York.
    Wang, Y., & Witten, I. (1997). Inducing Model Trees for Continuous Classes. Paper presented at the Proceedings of the Poster Papers of the European Conference on Machine Learning, University of Economics, Faculty of Informatics and Statistics, Prague.
    Wang, Y. F. (2003). On-demand forecasting of stock prices using a real-time predictor. IEEE Transactions on Knowledge and Data Engineering, 15(4), 1033-1037.
    Willemain, T. R., Bress, R. A., & Halleck, L. S. (2003). Enhanced simulation inference using bootstraps of historical inputs. IIE Transactions, 35(9), 851-862.
    Wolpert, D. H. (1992). Stacked Generalization. Neural Networks, 5(2), 241-259.

    無法下載圖示 校內:2014-12-09公開
    校外:不公開
    電子論文尚未授權公開,紙本請查館藏目錄
    QR CODE