本研究主要目的是結合資料包絡分析法與決策樹規則運算應用於基金選購模型建立，希望可以透過實證模型的研究了解結合兩運算法則於國外基金的適用性，並找出篩選基金的重要屬性，以提供投資人評估共同基金持有組合另一個簡單的決策法則。本研究的範疇為國外四大種類基金，基金績效指標的資料來源為富盈財務資料庫；此外，基金淨值則由台灣經濟新報資料庫取得，而前期利用資料包絡分析法獲得各基金效率值使用工具為Frontier Analyst；後期決策樹模型的建構及其後樣本外的模擬則使用WEKA軟體做為分析的工具。本研究共提出兩個模型：依總體經濟的不同可分為多頭模型及空頭模型。多頭模型與空頭模型的樣本期間為2006年6月到2009年2月，分為樣本內及樣本外兩部分，樣本內(訓練組)的部分進行模型的建構，而樣本外(測試組)的資料則用來測試模型分類的正確性及有效性。綜合兩個模型的實證結果，我們可以建議投資人若對於未來行情持樂觀態度或總體經濟反轉向上，可以採用多頭模型來篩選持有的投資組合；反之，若投資人看空市場或總體經濟破底向下，則建議使用空頭模型作為投資決策的準則，同時強調風險的控管與報酬的追求，都可以篩選出同類型中在未來表現較好的基金群組。透過本研究樣本外期間分類正確率及投資報酬率的模擬，顯示具顯著性的績效指標的確可以篩選出未來報酬率較好的基金群組，證實結合DEA與決策樹所建構的模型有一定的成效，值得進一步的研究與應用。

The purpose of this research is to build up a model for mutual fund selection by combining Data Envelopment Analysis (DEA) and Decision Tree. It is expected to understand the suitability for combining these two operation on oversea mutual fund through experiment model study, and then to offer another simple decision rule to investor for holding combination on mutual fund. The range in this research is limited to four kinds of the overseas mutual fund. The data base for the fund performance evaluation index is acquired from Folion Financial Technology Co., Ltd., and net value of mutual fund is adopted from Taiwan Economic Journal Co., Ltd. The first step is to gain the efficiency of fund by DEA with Frontier Analyst. The later period is to build up a model of decision tree with WEKA software. The research finally proposes two models, including bull model and bear model varied with macroeconomic situation. The sample period is from June/2006 to Feb/2009. The sample is composed of training data and test data. The train data can be built up the model, and the test data can be used to confirm the accuracy of model. This research suggests that investor can apply for the bull model about fund holding combination when the economic grows up in the future. Oppositely, the bear model is used as economic declines. These two models also emphasizes both risk control and profit aspiring. In addition, this research discovers the best funds from the inside of overseas mutual fund. The significance performance evaluation index can be used to discover the best funds from the inside of overseas mutual fund by the model simulation conducted in this research. The model of combining DEA with Decision Tree thus is effective in selecting fund.

參考文獻
中文文獻
[1] 王郁仁，「股票型基金綜合指標與基金績效動態監控」，國立中山大學財務管理研究所，2005年
[2] 吳靜婷，「共同基金績效持續性及投信公司績效差異之探討」，國立中央大學企業管理研究所，1998年
[3] 吳姿瑤，「國內開放式股票基金在分類與預測模型比較之研究」，國立成功大學統計所，2005年
[4] 林秋瑾，「運用資料包絡分析法與決策樹建構司法機關效率評估模式之研究」，國立中正大學會計與資訊科技所碩士論文，2007 年。.
[5] 周榮祥，「以資料探勘技術分析股市個人投資績效之研究」，世新大學資訊管理研究所碩士論文，2003年。
[6] 高強、黃旭男，「管理績效評估：資料包絡分析法」，台北：華泰文化事業公司，2003年。
[7] 徐銘甫，「整合財務比率與資料包絡分析法於企業危機診斷模式之建構－約略集合、支援向量機與決策樹之應用」，中國文化大學會計研究所，2007年。
[8] 徐靜瑩，「CART決策樹分群法於基金績效之研究」，國立中山大學財務管理研究所，2005年。
[9] 張志宏，「台灣共同基金投資績效評估之研究」，國立成功大學企業管理研究所，1996年。
[10] 陳暐中，「共同基金技術效率評估」，國立中正大學財務金融研究所，1998年。
[11] 陳建文，「醫院經營效率衡量模式建立-DEA、DEA-ANN、CART與Malmquist方法應用」，國立台北科技大學商業自動化與管理研究所，2007年。
[12] 鄭忠樑，「運用分類樹於股價報酬率之研究」，元智大學資訊管理研究所碩士論文，2002年。
[13] 蘇新仁，「以決策樹建構資料包絡法之績效評估模式」，國立雲林科技大學工業工程與管理研究所，2004年。
[14] 葉明威，「應用資料探勘技術分析股市漲跌型態之研究」，國立中山大學資訊管理研究所，2003年。
英文文獻
[1] Chevalier, J. and Ellison, G. (1999) ‘Are some mutual fund managers better than others? Cross-sectional patterns in behavior and performance’, Journal of Finance, Vol.56, pp.875-899.
[2] Charnes, A., Cooper, W. W. and Rhodes, E.(1981) ‘Evaluating program and managerial efficiency: An application of DEA to program follow through’, Management Science, vol. 27, No.6, pp.668-697.
[3] Eric, H.S., Keith L.M., and Cheep K.O.(2000) ‘The Decision Tree Approach to Stock Selection– An Evolving Model Performs the Best’, The Journal of Portfolio Management, pp. 42-52.
[4] Fan, J. and Wen, P.(2007) ‘Application of C4.5 algorithm in web-based learning assessment system’, Proceedings of the Sixth International Conference on Machine Learning and Cybernetics, Vol.7, pp.4139-4143.
[5] Farrell, M.J. (1957) ‘The Measurement of Productive Efficiency’, Journal of the Royal
Statistical Society, Vol.120, No.3, pp.253-281.
[6] Gladish, B.P., Jones, D.F., and Tamiz, A. (2007) ‘An interactive three-stage model for mutual funds portfolio selection’, Omega - International Journal of Management Science, Vol.35, No.1, pp.75-88.
[7] Hubner, G (2007) ‘How do performance measures perform?’, The Journal of Portfolio Management, Vol.33, No.4, pp.64-74.
[8] Jensen, M.C. (1968) ‘The performance of mutual funds in the period 1945-64’, Journal of Finance, Vol.23, No.2, pp.389-416.
[9] Keith, C., Dirk, N. and Niall, O. (2007) ‘UK mutual fund performance: Skill or luck?’, Journal of Empirical Finance, Vol.15, pp.613-634.
[10] Kosowski, Robert, Timmermann, Allan, White, Hal, Wermers, and Russ, (2006) ‘Can mutual fund “stars” really pick stocks? New evidence from a bootstrap analysis’, Journal of Finance LXI, Vol.6, pp.2551–2595.
[11] Lee, C.F. and Rahman, S. (1991) ‘New evidence on timing and security selection skill of mutual fund mangers’, The Journal of Portfolio Management, Winter, pp.80-83.
[12] Michael J.A. Berry and Gordon S. Linoff, (1997) ‘Data Mining Techniques：for Marketing, Sales, and Customer Support‘, John Wiley and Sons, Inc., New York.
[13] Samoilenko S., Osei-Bryson KM. (2008) ‘Increasing the discriminatory power of DEA in the presence of the sample heterogeneity with cluster analysis and decision trees’, Expert Systems with Applications, Vol.34, No.2, pp.1568-1581.
[14] Seol, H., Choi, J., Park, G., and park, Y. (2007) ‘A framework for benchmarking service process using data envelopment analysis and decision tree’, Expert Systems with Applications, Vol.32, No.2, pp.432-440.
[15] Sorsensen, E.H., Miller, K. L., and Ooi, C.K. (2000) ‘The decision tree approach to stock selection – An evolving model performs the best’, The Journal of Portfolio Management, Vol.27, No.1, pp.42-52
[16] Sohn, S. Y., and T. H. Moon. (2004) ‘Decision Tree based on data envelopment analysis for effective technology commercialization’, Expert Systems with Applications. Vol.26, No.2, pp.279-284.
[17] Sohn, S. Y., and T. H. Moon. (2005 Fall) ‘Intelligent approach for effective management of governmental funds for small and medium enterprises’, Expert Systems with Applications. Vol.29, No.3, pp.566-572.
[18] Tokunaga, H., Atlam, E. S., Fuketa, M., Morita, F., Tsuda, K. and Aoe, J.I. (2004) ‘Estimating sentence types in computer related new product bulletins using a decision tree’, Information Science, Vol.168, pp.185-200.
[19] Yun, Y. B., Nakayama, H and Arakawa, M. (2004) ‘Multiple criteria decision making with generalized DEA and an aspiration level method’, European Journal of Operational Research, Vol.158, pp.697-706.