隨著資訊進步，雲端技術的開發，大量的數據資料充斥在其中，迄今仍有許多未知領域資料仍放置在某個角落等待被挖掘，而資料探勘(Data Mining)與機器學習(Machine Learning)的技術演進，帶給我們更多更新的技術來從資料中萃取出對人們於生活上、知識上、研究上更多有用的知識。然而分類是資料探勘中滿重要的一環技術。利用已知資料與分類屬性，來建立資料分類的模型，透過模型的建立來預測未知新資料。支持向量機(Support Vector Machine, SVM)是一種以統計理論作為基礎並在近年被廣泛使用的分類方式。藉由已知資料與資料分類屬性，來建立出一個分類模型。此分類器雖可從大量的數據資料中透過資料屬性將資料映射至高維空間來將資料分類出，但是，仍有許多矛盾的資料存在，導致無法正確地分類出來。本研究理論是藉由模糊理論將可能存在於資料集中的矛盾資料找出並且消除，以重新建立一組新的訓練模型來作為新資料的預測。為了驗證理論，由UCI(University California Irvine)的資料庫中，找尋三個存在於真實世界中分類問題的資料，將這些資料透過模糊理論去除造成分類效率不佳的資料點，能有效地提升新資料的分類準確率。

With advent of information and cloud technology development, there are still numerous unknown data that could be used for research and waiting to be excavated. The evolution of Data Mining and Machine Learning technology has brought us even newer front to extract useful information to benefit people’s daily lives, general knowledge, and research applications. However, the Data Mining Classification is an important part of technology. By using the known data and categorizing properties to build a classification model, we could apply this model to further predict the unknown new data. Support Vector Machine (SVM) is a type of statistical theory widely used in recent years and based on the data attribute to build a classification model. This classifier can remap the data to high-dimension space through the attribute of information from a large amount of data. Nevertheless, there still exists conflicting information preventing the data to be accurately classified. In this thesis, we attempt to eliminate the data that may be inconsistent in dataset by applying fuzzy theory first, and followed by re-establishing a new training model for prediction. In order to verify this theory, three existing real world categorizing dataset problems were drawn from the UCI database first, followed by the removal of outlier data by the use of fuzzy theory, and finally achieving a substantially greatly improved classification accuracy for new data.

[1] V. N. Vapnik, “The Nature of Statistical Learning Theory, ” Springer-Verlag, Berlin Heidelberg, New York. 1995.
[2] V.N. Vapnik, “An Overview of Statistical Learning Theory,” IEEE Transaction on Neural Networks, vol. 10, pp 988-999, 1999.
[3] C. F. Lin and S. D. Wang, “Fuzzy support vector machines,” IEEE Transaction on Neural Networks, vol. 13, pp 467-471, 2002.
[4] A.Shiton, M.Palaniswami, D.Ralph and A.C.Tsoi, “Incremental training of support vector machines,” IEEE Transactions on Neural Networks, vol.16, pp 114-131, 2005
[5] J.S.Taylor, “An Extension on Statistical Comparisons of Classifiers over Multiple Data Sets for all Pairwise Comparisons,” Journal of Machine Learning Research 9, pp. 2677-2694, Aug. 2008.
[6] Chih-Jen Lin, Chih-Chung Chang, Chih-wei Hsu, “A Pratical Guide to Support Vector Classification,” National Taiwan University, Department of Computer Science. April 15, 2010.
[7] Md Omar Faruqe, Md Al Mehedi Hasan, “Face Recongition Using PCA and SVM,” Anti-counterfeiting, Security, and Identification in Communication, 2009. ASID 2009. 3rd International Conference, pp.97-101, 2009.
[8] C.W Huang, “Analysis of Breast Cancer Using Meridian Energy,” National Chung Cheng University Computer Science and Information Engineering.2008.
[9] Shan Meng, Jingxiong Huang, Youwei Zhang,”An SVM/PCA-Based Method of Human Face ROI Localization,” Signal Processing, 2004. Proceedings. ICSP '04. 2004 7th International Conference on Vol.2. pp.963-966, 2004.
[10] Nello Cristianini and John Shawe-Tylor, “An Introduction to Support Vector Machines and other Kernel-based Learning Methods”.3rd Edition, 2003, Cambridge University Press.
[11] Eric C.C.Tsang, Daniels. Yeung, and Patrick P.K Chan “Fuzzy Support Vector Machines For Solving two-class Problems,” Machine Learning and Cybernetics, 2003 International Conference on. Vol.2. pp.1080-1083. 2003.
[12] T.Inoue and S.Abe “Fuzzy Support Vector Machine for Pattern Classification,” Neural Networks, 2001. Proceedings. IJCNN '01. International Joint Conference on. Vol.2. pp. 1449-1454. 2001
[13] Chen Chi Lin, “Real-Time face gender recognition system based on average face difference mask and features from 2D-DCT,” National Taiwan University of Science and Technology Department of Electrical Engineering.
[14] Xiaoyu Zhang, “Face detection based on Two Dimensional Principal Component Analysis and Support Vector Machine,” Mechatronics and Automation, Proceedings of the 2006 IEEE International Conference on.pp.1488-1492. 2006
[15] Ying-Wen Bai, “Using Fuzzy Logic to Reduce Power Consumption of Notebooks,” Consumer Electronics, 2009. ISCE '09. IEEE 13th International Symposium on. pp.488-492. 2009.
[16] Ying-Wen Bai, “Using an Embedded Controller with Fuzzy Logic to reduce power consumption of Mobile Computers,” IECON 2010 - 36th Annual Conference on IEEE Industrial Electronics Society. pp.236-241. 2010
[17] P.M. Murphy, UCI-Benchmark Repository of Artificial and Real Data Sets, http://www.ics.uci.edu/~mlearn, University of California Irvine, CA, 1995
[18] Steven D. Kaehler, “Fuzzy Logic Introduction,” http://www.seattlerobotics.org/encoder/mar98/fuz/fl_part1.html, Seattle Robotics Society.
[19] http://en.wikipedia.org/wiki/Lagrange_multiplier
[20] W. Karush. “Minima of Functions of Several Variables with Inequalities as Side Constraints. M.Sc. Dissertation,” Dept. of Mathematics, Univ. of Chicago, Chicago, Illinois. 1939.
[21] H.W. Kuhn and A.W. Tucker, “Nonlinear programming,” Proceedings of 2nd Berkeley Symposium. Berkeley University of California Press. pp. 481-492. 1951.
[22] L.A.Zadeh, “Fuzzy sets,” Information and Control, Vol.8, No.1, pp. 338-353. 1965.
[23] W.C.Chen, “The use of Fuzzy mathematical programming in Support Vector Machines,”Master Thesis, Department of Electrical Engineering National Central Universary, 2010.
[24] M.F.Han, “Fuzzy Support Vector Machines with Uncertainty of Parameter C,”Master Thesis, Department of Electrical Engineering National Central Universary, 2008.