馬氏田口系統(Mahalanobis-Taguchi System, MTS)是田口玄一博士結合田口法(Taguchi method)與馬氏距離(Mahalanobis distance function)，所提出作為衡量多維度系統的一個高績效且具有成本效益的方法論。目前已廣泛應用於各產業中，且均有良好的成效。然而，在營建產業方面，仍然缺乏此方法之應用。此外，文獻顯示MTS在運作上仍有某些問題點。
藉此，本研究擬藉由MTS建立一套系統診斷模型，透過模糊理論改進MTS對於非比例尺度資料處理上的問題點，並採用Penny所提出之馬氏距離的閥值運算方法，改善閥值訂定的效率。最後，本研究將此模型應用於邊坡道路的崩塌診斷，針對颱風豪雨期間所導致之邊坡崩塌案例進行研究，建構山區道路邊坡崩塌潛能診斷模型及邊坡崩塌類型診斷模型。此外，本研究更進一步將MTS與傳統鑑別分析進行比較，結果顯示MTS具有較好的成效。
本研究所建構的系統診斷模型，可供工程管理或研究人員參考，進一步應用於營建業之相關議題，藉此即時診斷系統的狀態與異常程度，以提高績效及成本之效益。

Mahalanobis-Taguchi System (MTS), developed by Dr. Taguchi Genichi, combines Taguchi methods and Mahalanobis distance function to measure the situation of multi-dimensional system, which is a cost-effective and high performance methodology. Currently, MTS has been comprehensively employed in various fields with satisfied applications. However, MTS still has certain problem in practice and seldom applications in the area of construction industry.
Therefore, this study adopts the methodology of MTS to establish a system diagnosis model. Also, this work makes an improvement for MTS on the data-processing issue of non-scale measurement by utilizing Fuzzy Theory. Then, the threshold calculation function developed by Penny is adopted to improve the efficiency of determining the threshold. Finally, this study adopts the case of roadside mudslide diagnosis to demonstrate the feasibility and practicability of the proposed model. Furthermore, this study also compares the performance between MTS and the traditional Discriminant Analysis. Results show that MTS has better performance on this case.
The system diagnosis model constructed in this research is useful for engineering industry or research staffs and can be applied to relevant issues in construction industry for instant diagnosis of system status to improve the performance and cost effectiveness.

1.Taguchi, G. and R. Jugulum, “The Mahalanobis-Taguchi Strategy”, John Wiley and Sons Press, New York, 2002.
2.吳明隆，「SPSS與統計應用分析」，五南圖書出版股份有限公司，2005年7月。
3.Michael H. Kutner, Christopher J. Nachtseim, and John Neter，「應用線性迴歸模型」，華泰文化事業股份有限公司，2005年7月。
4.Hberty C. J. , “Problems with stepwise methods-Better alter-natives”, In B. Thompson (Ed.), Advances in Social Science Methodology, Vol.1, pp.43-70,1989.
5.Thompson B., “Common methodology mistakes in dissertations: Improving dissertation quality” ,Paper presented at the annual meeting of the Mid-South Educational Research Association, Louisville.,ERIC Document Reproduction Service No. ED 301 595,1988.
6.葉怡成，「應用類神經網路」，儒林圖書有限公司，2002年6月。
7.莊榮仁，「馬氏-田口系統與類神經網路的比較」，化工技術第11卷第2期，2003年2月。
8.吳復強，「產品穩健設計:田口方法之原理與應用」，高立圖書有限公司，2005年7月。
9.Taguchi, G., S. Chowdury, and Wu, Y., “The Mahalanobis-Taguchi System”, McGraw-Hill Press,New York ,2001.
10.Kanetaka, Tatsuji., “Diagnosis of a Special Health Cjeck Using Mahalanobis Distance ”, ASI Journal,Vol.3,No.1,1990.
11.Taguchi, Genichi. “Application of Mahalanobis Distance for Medical Treatment.” Journal of Quality Engineering Forum,Vol.2,No.6.
12.Asada, Masashi, “Yield Prediction by Using Mahalanobis Distance”, Quality Engineering Forum Symposium,1998.
13.Lavallee, Louis.,“Application of Mahalanobis Taguchi System to Thermal Inkjet Image Quality Inspection” ,ASI symposium, 1998.
14.Mizoguchi, Kazutaka, Akiko Kawai, Kazutaka Hamada(Nissan Motor Co.),Takashi Kamoshita(National Research Laboratory of Metrology).”A Research on a Sensing System Using Mahalanobis Distance for Prevention of Driving Accidents,” Quality Engineering Forum Symposium, 1998.
15.許前、鄭稱德、韓之俊，「MTS多類判別研究」，Journal of Nanjing University of Science and Technology,Vol.26 No.1,2002.
16.Taketoshi Riho, Akihito Suzuki, Junji Oro, Kenji Ohmi, and Hajime Tanaka, “The Yield Enhancement Methodology for Invisible Defects Using the MTS+ Method”, IEEE TRANSACTIONS ON SEMICONDUCTOR MANUFACTURING, Vol. 18, No. 4, November 2005.
17.Makabe, Akira, Kei Takada(Seiko Epson Corp) Hiroshi Yano (Japanese Standards Association). “An Application of the Mahalanobis Distance for Solder Joint Appearance Automatic Inspection,” Quality Engineering Forum Symposium, 1998.
18.Kamoshita, Takashi, Kazuta Tabata, Harutoshi Okano, Kazuhito Takahahi, Hiroshi Yano., “Application of Mahalanobis Taguchi System to the Fault Diagnosis Program”, ASI Total Product Development Symposium,1997.
19.游士輝，「MTS在測試流程改善的應用--以筆記型電腦為例」，國立交通大學碩士論文，2002年。
20.蕭宇翔，「應用MTS於非平衡資料分析之穩健性研究 - 以行動電話檢測流程為例」，國立交通大學碩士論文，2004年。
21.廖合，「應用資料包絡法降低電源轉換器溫升之研究」，國立政治大學碩士論文，2005年。
22.程紀嘉，「具多重反應值製程之要因分析」，國立成功大學碩士論文，2005年。
23.Kanazana, Yukihiko, Shinzou Kishimoto, Jun Okamoto(Fuji Photo Film Co). “An Application of MTS for Qualitative Diagnosis of Photographic Processing Solution” ,Quality Engineering Forum Symposium, 1998.
24.Asahara, Hatsuki, Yuichiro Ota, Shuichi Takeda, Masahiko Marumoto, Toshitsugu Sato(TsumuA & Co.).”Pattern Recognition for Infrared Absorption of Spectrum” ,Quality Engineering Forum Symposium, 1998.
25.Prasun Das, Shubhabrata Datta, “Exploring the effects of chemical composition in hot rolled steel product using Mahalanobis distance scale under Mahalanobis–Taguchi system”, Computational Materials Science,2006.
26.Matsuda, Rika, Yoshiki Ikeda (Mitsubishi Space Software Co.)Takashi Kamoshita(National Research Laboratory of Metrology), Kazyuyuki Touhara(National Space Development Agency of Japan)., “Application of Mahalanobis Taguchi System to the Fault Diagnosis Program” ,ASI Symposium,1998.
27.Kamoshita,Takashi, Ken-ichi Okumura, Kazuhito Takahashi, Masao Mesumura, Hiroshi Yano., “Character Recognition Using Mahalanobis Distance” ,Quality Engineering Forum Symposium,1997.
28.Takada, Kei(Seiko Epson Corp.), Kazuhito Takahashi(University of Electro-Communication),Hiroshi Yano (Ohken Associates), “The Study of Valuation of Programmer’s Capability by MTS Method” , Quality Engineering Forum Symposium,1998.
29.Sakano, Susumu ,“U.S. Dollar Bill Inspection” ,Quality Engineering Forum Symposium,1998.
30.Mitsuyoshi Nagao, Masahito Yamamoto, Keiji Suzuki and Azuma Ohuchi, “MTS Approach to Facial Image Recognition”,Research Group of Complex Systems Engineering,Graduate School of Engineering, Hokkaido University,1999.
31.陳美芳，「結合適性診斷測驗與馬氏距離篩選適任的檢查員」，私立逢甲大學碩士論文，2005年。
32.Hirohisa Aman, Member, Naomi Mochiduki, Nonmember, and Hiroyuki Yamada, Member,“A Model for Detecting Cost-Prone Classes Based on Mahalanobis-Taguchi Method”, IEICE TRANS. INF. & SYST., Vol.E89–D, No.4 April 2006.
33.王慧君，「馬氏-田口系統之特性探討及其於信用評分之應用」，國立交通大學碩士論文，2003年。
34.薛跃、韓之俊、盛黨紅，”MTS法用於上市公司財務品質評估初探”，中文核心期刊，24卷1期，2005年。
35.Hiroshi Morita and Yosuke Haba, “Variable Selection in data envelopment analysis based on external evaluation”, Variable Selection in DEA
36.蘇朝墩, “A mahalanobis distance based classifier for diagnosis of diseases”, journal of the Chinese institute of industrial enguneers,Vol.19,no.5,pp.41-47,2002.
37.Woodall, W. H., R. Koudelik, K. L. Tsui, S. B. Kim, Z. G. Stoumbos, and C. P. Carvounis, “A review and analysis of the Mahalanobis-Taguchi system” ,Technometrics, February, 45, 1-15,2003.
38.葉坤海，「模糊判別分析結合MTS之績效評估」，國立雲林科技大學碩士論文，2005年。
39.王凱仁，「建設公司財務危機動態預警模型之研究」，國立交通大學碩士論文，2002年。
40.Zadeh, L.A., “Fuzzy Set” , Information and Control, No. 8,pp.338-353 ,1965.
41.Kaufmann, A. and Gupta, M. M., “Introduction to fuzzy arithmetic theory and applications”, New York: van Nostrand Reinhold, 1991.
42.Klir, G. J. and Yuan, B., “Fuzzy sets and fuzzy logic：theory and applications”, Prentice Hall PTR, New Jersey,1995.
43.Zadeh, L. A., “The concept of a linguistic variable and its application to approximate reasoning”, Information Sciences, 8,199-249(Ⅰ), 301-357(Ⅱ), 1975.
44.Zimmermann, H. J. , “Fuzzy set theory and its applications”, second edition, Kluwer Academic Publishers, Boston/ London.,1991.
45.Bellman, R. E. and Zadeh, L. A., “Decision-making in a fuzzy environment, Management Science”, 17:4, .141-164,1970.
46.Zimmermann, H. J. “Fuzzy set, decision making and expert system”, Kluwer, Boston,1987.
47.Buckley, J. J., “Ranking Alternatives Using Fuzzy Numbers, Fuzzy Sets and Systems”, 15(1)：21-31., 1985.
48Chen, S. J., Hwang, C. L. and Hwang, F. P., “Fuzzy multiple attributes decision making methods and Applications”, Springer-Verlag,1992.
49.公開資訊觀測站http://newmops.tse.com.tw/
50.台灣經濟新報資料庫之「危機大世紀」資料庫
51.Wilk S.S, “Multivariate Statistical Outlier”,Sankhya A,25, 407-426.,1963.
52.Kay I. Penny , “Appropriate Critical Values When Testing for a Single Multivariate Outlier by Using the Mahalanobis Distance “,Vol. 45, No. 1 pp. 73-81,1996.
53.趙衛君，「應用高斯過程建立分階式山區道路邊坡崩塌預測模式之研究」，國立台灣科技大學碩士論文，2004年。
54.Hoek, E. and Bray, J., “Rock Slope Engineering”, Institution of Mining and Metallurgy,1977.
55.林建煌，「策略組群、組織作為與績效關係研究-飲料業實證研究」，國立政治大學博士論文，1989年。