臺灣博碩士論文加值系統

都市計劃的決策者在制定決策時，常準備不同的評估方案來做為選擇與採用，而運用各種的決策理論與方法以作為評估方案。
究竟何種決策理論的方法較能讓決策者做出較符合公平正義原則的決策，實證和理論上仍為模糊的概念。許多研究都指出，不同偏好結構屬性的強度會影響到決策結果，例如：(Keeny ＆ Raiffa，(1976)多屬性效用理論；Dyer ＆ Sarin，(1979)可衡量加法價值函數等。而Lai ＆ Hopkins，(1995)曾在決策者不能穩定表示屬性得失的偏好強度。因此，如何增進偏好強度衡量的理論研究，是決策理論的探討上非常重要的一個環節。
本研究目的；在理論上以界定影響決策方案的屬性中，偏好強度的意義及探討，在實驗設計上嘗試應用類神經網路在決策者制定決策與個別決策者主觀決策比較性，比較人(受測者)與機(類神經網路)的決策模式，探討人機比較在決策測定品質上何種方式品質較高，並且也判斷電腦輔助的決策與人決策的差異性，以交替式的人機實驗設計的方式，來比較兩種實驗方式的效率性。
本研究從實驗的結果顯示，人(受測者)與機(類神經網路)在決策方案屬性中為機(類神經網路)的實驗方式品質來的高，加入交替式實驗程序的分析結果顯示出依然是機的測定品質較高，然而決策方案屬性間人與機的比較依然是機(類神經網路)的實驗方式品質高，但是在統計分析上，差距卻不明顯，在探討男女對於決策上的比較，發現結果差距不明顯，另外在比較加入交替式實驗程序的前後測定品質，發現交替式實驗可提高測定品質；另外以決策方案屬性中與屬性間偏好結構分析結果發現，以決策方案屬性中加入交替式機(類神經網路)與屬性間權重加入交替式之機(類神經網路)實驗方式品質較高。

Policy makers usually select and adopt diverse evaluating proposals while making policies on urban development. They evaluate proposals by the application of the received policy analyses. Which is the best method for policy makers to make the policies in terms of just principles? In theoretical and empirical studies, the concept of strength of preference is still an elusive notion. However, many theories indicate that different structures of strength of preferences affect the result during the process of policy decision, i. e., multi-attribute utility theory developed by Keeny& Raiffa and measurable additive value theory (MAVT), which was indicated by Dyer＆Sarin. However, Lai and Hopkins indicated that policy makers cannot make stable judgments on strength of preference in 1995. Therefore, it is a crucial link to enhance the research on measurement of strength of preference in empirical studies.
The purpose of the research is to discuss the significance of the concept of strength of preference to define the influencing decision scheme of attributes in theoretical studies. The research intends to apply neural network in an experimental design to formulate the comparison, based on the policy maker and the individual policy-maker making policies subjectively, between the decision model of the human subjects and the machine (neural network), and we intend to explore in which model the decision quality is higher in making policies. In addition, we judge the difference between computer aided decision-making and human decision-making by an iterative procedure to compare the effectiveness of the two experimental modes.
The experimental results showed that the decision quality is higher for the machine than the human subjects. In particular, the decision quality was higher for the machine that the human subjects, even in the human/machine iterative procedure, both with and between attributes. The statistical analysis showed the disparity was not significant in comparing the female and male human subjects, that the iterative procedure could increase the validity of the experiment, and that the machine outperformed the human subjects when the preference structures were taken into account including the weights across the attributes.

第一章 緒論 ........................................ 1
第一節 研究動機與目的的 ............................ 1
第二節 研究範圍與方法 .............................. 4
第三節 研究內容與流程 .............................. 6
第二章 文獻回顧與探討 .............................. 8
第一節 多屬性效用理論及偏好強度研究 ................ 8
第二節 可衡量加法價值函數對等偏好判斷研究...........13
第三節 人工智慧相關文獻探討 ....................... 17
第四節 類神經網路及倒傳遞網路探討及相關研究 ........19
第三章 研究設計 ................................... 24
第一節 實驗材料選擇 ................................25
第二節 實驗設計 ....................................27
第三節 人(受測者)問卷設計 ..........................32
第四節 機(類神經網路)實驗設計 ......................37
第五節 人機交替式實驗設計 ..........................39
第六節 評估準則 ....................................40
第七節 研究限制 ....................................43
第四章 實驗過程及結果 ..............................44
第一節 實驗過程 ....................................44
第二節 實驗結果分析 ................................47
第五章 結論與建議 ..................................60
第一節 結論 ........................................60
第二節 建議與後續研究 ..............................61
第六章 參考文獻 ....................................62
附錄................................................69

中文部分：
��李士，(1985)，「利用多屬性模式對消費者偏好之研究」，東海大學企業管理研究所碩士論文。
��賴世剛，(1994)，「多屬性決策理論在都市計劃之應用」，人與地，第130期。
��許仁成，(1995)，「多屬性決策方法評估標準之研究-特徵向量法的應用」，國立中興大學都市計劃研究所碩士論文。
��辜永奇，(1995)，「多屬性決策方法中偏好強度的意義與測定」，國立中興大學都市計劃研究所碩士論文。
��賴世剛、辜永奇，(1997)，「等價與比率偏好強度判斷之實驗比較」，管理與系統，第四卷，第二期，75-90。
��蘇木春、張孝德，(1999)，機械學習：類神經網路、模糊系統以及基因演算法則，全華科技圖書股份有限公司。
��葉怡成，(2003)，應用類神經網路，儒林圖書公司。
��葉怡成，(2003)，類神經網路模式 應用與實作，儒林圖書公司。
��陳耀茂、殷純淵，(2004)，類神經網路 PCNeuron。
��廖敏伶，(2005)，「應用類神經網路於果蔬汁之生產預測」，僑光技術學院管理研究所碩士論文。
��李城忠，(2008)，應用統計學SPSS完全攻略，新文京圖書公司。

西文部分：
��Bell,D.V.,Raiffa,H.,and Tversky A.(1988),Decision Making,New York：Cambridge University Press.
��Dyer,J,S.,and Sarin ,R.K.，(1979)，“Measurable Multiattribute Value Functions”，Operation Research，27(4)：810-822.
��Dyer,J,S.,and Sarin ,R.K.，(1980)，“Group Perfernce Aggregaation Rules Based on Preference，“Management Science 25(9)：822-832.
��Hwang, C. L. and Yoon, K. Multiple Attribute Decision Making Methods and
Applications, Springer, Berlin Heidelberg, 1981.
��Kahneman, D. & Tversky, A. (1979). Prospect theory: An analysis of decisions under risk. Econometrica, 47,263-291.
��Keller L.R.(1985),“An Empirical Investigation of Relative Risk Aversion”,IEEE Transactions on System, Man ,and Cybernetics SMC-15(4)：475-482.
��Keeney R.L.,and Raiffa H.(1976),Decision with Multiple Objectives,New York：John Wiley.
��Keeney, R. L. & Raiffa, H. (1993). Decisions with multiple objectives. New York:Cambridge University Press.
��Lai, Shih-Kung, Hopkins, Lewis D. (1989), “The meanings of trade-offs in multiattribute evaluation methods: a comparison,” Environment and Planning B: Planning and Design, 16, 155-170.
��Lai, Shih-Kung and Hopkins, Lewis D., (1990), Ability to express tradeoff judgments in multi-attribute utility theory and the Analytic Hierarchy Process, Planning Papers, Department of Urban and Regional Planning, University of Illinois at Urbana-Champaign.
��Lai, Shih-Kung, (1995), “A preference-based interpretation of AHP,” Omega, 23(4), 453-462.
��Satty T.L.(1980),The Analytic Hierarchy Process, New York：McGraw-Hill.
��Sarin R.K.(1982),“Strength of Preference and Risky Choice ”Operations Research,30(5)：982-997.