論文使用權限 Thesis access permission:校內校外完全公開 unrestricted
開放時間 Available:
校內 Campus: 已公開 available
校外 Off-campus: 已公開 available
論文名稱 Title |
使用輻射基底函數網路建構多屬性效用函數 Using Radial Basis Function Networks to Model Multi-attribute Utility Functions |
||
系所名稱 Department |
|||
畢業學年期 Year, semester |
語文別 Language |
||
學位類別 Degree |
頁數 Number of pages |
67 |
|
研究生 Author |
|||
指導教授 Advisor |
|||
召集委員 Convenor |
|||
口試委員 Advisory Committee |
|||
口試日期 Date of Exam |
2004-07-07 |
繳交日期 Date of Submission |
2004-07-14 |
關鍵字 Keywords |
決策分析、效用函數、輻射基底函數、多屬性效用理論、簡易多屬性評等技術 MAUT, SMARTS, Decision Analysis, SMARTER, Utility Function, RBF |
||
統計 Statistics |
本論文已被瀏覽 5713 次,被下載 3028 次 The thesis/dissertation has been browsed 5713 times, has been downloaded 3028 times. |
中文摘要 |
隨著電子商務的蓬勃發展,許多線上協商、議價系統必須依賴決策分析領域中的效用函數協助使用者找出可行解決方案,或者是對使用者效用最大的方案。因此,如何快速、正確地建構使用者效用函數,成為一個重要的課題。本研究提出人工智慧中的輻射基底函數類神經網路方法,期能更為快速、精確地建構使用者的效用函數。透過實驗驗證此方法的可行性,筆者分別將輻射基底函數、複回歸、簡易多屬性評等技術等建構效用函數的方法,就「預測效能」、「花費時間」以及「使用者對這些方法的認知」作比較。實驗結果證實本研究所提出的輻射基底網路方法可行,不僅比過去決策分析所常用的簡易多屬性評等技術省時,同時也較複回歸、簡易多屬性評等技術等方法精確。 |
Abstract |
On-line negotiation and bargaining systems can work effectively on the Internet based on the prerequisite that user utility functions are known while undergoing transactions. However, this prerequisite is hard to meet due to the variety and anonymous nature of Internet surfing. Therefore, how to rapidly and precisely construct a user’s utility function is an essential issue. This research proposes a radial basis function (RBF) network, a neural network, to model a user’s utility function in order to rapidly and precisely model user utility function. We verify the feasibility of the method through experiments, and compare the performance of RBF networks in prediction performance, time expenses, and subjects’ perceptions with the Multiple Regression (MR), SMARTS, and SMARTER methods. The results show that the RBF network method is feasible in these criteria. Not only the RBF network needs less time to construct the users’ utility function than the SMARTS method does, but also it can model user utility functions more precisely than the MR, SMARTS, and SMARTER methods. |
目次 Table of Contents |
中文摘要 III ABSTRACT IV TABLE OF CONTENTS V LIST OF TABLES VII LIST OF FIGURES VIII CHAPTER 1 INTRODUCTION 1 1.1 RESEARCH BACKGROUND 1 1.2 RESEARCH MOTIVATION 1 1.3 RESEARCH OBJECTIVES 2 CHAPTER 2 LITERATURE REVIEW 3 2.1 MULTI-ATTRIBUTE UTILITY THEORY 3 2.1.1 Attributes 3 2.1.2 Single-dimension Utility Functions 4 2.1.3 Model Types 4 2.1.4 Weights 5 2.2 HOW TO MODEL A MULTI-ATTRIBUTE UTILITY 6 2.3 RADIAL BASIS FUNCTION NETWORK 9 2.3.1 Framework of Radial Basis Function Network 9 2.3.2 Training Algorithm 11 CHAPTER 3 USING RBF NETWORKS TO MODEL MAU FUNCTIONS 15 3.1 WHY TO USE RBF NETWORK TO MODEL MAU FUNCTIONS 15 3.2 RBF NETWORK LEARNING PROCEDURE TO MODEL MAU FUNCTIONS 16 CHAPTER 4 EXPERIMENTAL DESIGN 21 4.1 DECISION PROBLEMS 21 4.2 SUBJECTS 22 4.3 EXPERIMENTAL PROCEDURE 24 4.4 MODEL CONSTRUCTION 26 CHAPTER 5 EXPERIMENTAL RESULTS 28 5.1 PREDICTION PERFORMANCE 28 5.1.1 How well to predict a subject’s ordinal preferences 28 5.1.2 How well to predict interval preferences 31 5.1.3 Prediction Errors 33 5.1.4 Whether the construction treatments change subjects preferences 35 5.3 SUBJECT PERCEPTIONS TO RBF AND SMARTS 40 5.4 DISCUSSIONS 41 CHAPTER 6 CONCLUSION AND FURTHER WORK 44 6.1 CONCLUSIONS 44 6.2 FURTHER WORK 45 APPENDIX A 49 THE INSTRUMENTS USED FOR JOB SEARCH DECISION PROBLEMS IN EXPERIMENTS 49 A.1 The description of job search decision problem 49 A.2 Utility Function Construction using SMARTS 50 A.3 Utility Function Construction for RBF Network 54 A.4 Pre-test Benchmark Questions 55 A.5 Post-test Benchmark Questions 56 APPENDIX B 57 THE INSTRUMENTS USED FOR APARTMENT RENTING DECISION PROBLEMS IN EXPERIMENTS 57 B.1 The description of job search decision problem 57 B.2 Utility Function Construction using SMARTS 58 B.3 Utility Function Construction for RBF Network 64 B.4 Pre-test Benchmark Questions 65 B.5 Post-test Benchmark Questions 66 APPENDIX C 67 C.1 QUESTIONNAIRE USED FOR ASKING SUBJECTS’ PERSPECTIVES TOWARD MAU FUNCTION CONSTRUCTION PROCESS 67 |
參考文獻 References |
Broomhead, D.S., and Lowe, D. (1988). Multivariable functional interpolation and adaptive networks. Complex System, 2, 321-355. Bigus, J.P. (1996). Data mining with neural network. MacGraw-Hill Chen, S., Cowan, C.N., and Grant, P. M. (1991). Orthogonal least squares learning algorithm for radial basis function networks. IEEE Transactions on Neural Networks, 2 (2), 302-309. Clemen, R.T. (1996). Making Hard Decision: An Introduction to Decision Analysis. Belmont, CA: Duxbury Press. Dyer, J.S., and Sarin R.S. (1979). Measurable Multiattribute Value Functions. Operation Research, 27 (4), 810-822 Edwards, W., and Barron, H.F. (1994). SMARTS and SMARTER: improved simple methods for multiattribute utility measurement. Organizational Behavior and Human Decision Processes, 60, 306-325. Fischer, G.W. (1977). Convergent Validation of Decomposed Multi-Attribute Utility Assessment Procedures for Risky and Riskless Decisions. Organization Behavior and Human Performance, 18, 293-315 Fischer, G.W. (1979). Utility for multiple objectives decisions: Do they accurately represent human preferences? Decision Science, 10 (3), 451-471 Goodwin, P., and Wright G. (1998). Decision analysis for management judgement. John Willey and Sons. Haykin, S. (1994). Neural networks, a comprehensive foundation. Macmillan. Keeney, R., and Raiffa, H. (1976). Decisions with multiple objectives: preferences and value tradeoffs. John Willey and Sons. Ku, Y.C. (1994). The meaning and measurement of strength of preference in multiattribute decision making. Unpublished master thesis, National Chung Hsing University, Taiwan, Republic of China. Lin C.L. (2001). Improving the generalization capability of the RBF neural networks via the use of linear regression techniques. Unpublished master thesis, National Sun Yat-sen University, Taiwan, Republic of China. Lin, J.T., Bhattacharyya, D., and Kecman V. (2003). Multiple regression and neural networks analyses in composites machining. Composites Sciences and Technology, 63, 539-548. Liu, T.Z. (1996). On the improvements of regression and classification neural networks and their applications in nonlinear model control problems. Unpublished doctoral dissertation, National Sun Yat-sen University, Taiwan, Republic of China. Moody, J., and Darken, C. (1989). Fast-learning in networks of locally-turned processing units. Neural Computation, 1, 281-294. Russel, S., and Norvig, P. (2003). Artificial Intelligence: A Modern Approach (second edition). Englewood Cliffs, NJ: Prentice Hall. Sandholm, T. (1999). Automated negotiation. Communication of the ACM, 42 (3), 84-85. Schoemaker, P.H., and Waid, C.C. (1982). An experiment comparison of different approaches to determining weights in additive utility models. Management Science, 28 (2), 182-195 Siegal, S., and Castellan, N.J. (1988). Nonparametric Statistics for the behavioral sciences (2nd ed.). McGraw-Hill. Srivastava, J., Connolly T., and Lee R.B. (1995). Do Rank Suffice? A Comparison of Alternative weighting Approaches in Value Elicitation. Organization Behavior and Human Performance, 63, 112-116 von Winterfeldt, D., and Edwards, W. (1986). Decision analysis and behavior research. New York: Cambridge University Press. Wang, M.S., and Yang, J.S. (1998). A multi-criterion experimental comparison of three multi-attribute weight measurement methods. Journal of Multi-Criterion Decision Analysis, 7, 340-350. Yap, C.S., Raman, K.S., and Leong C.M. (1992, January). Methods for information system project selection: an experiment study of AHP and SMART. Paper presented in the Twenty-Fifth International Conference on System Sciences, Hawaii. |
電子全文 Fulltext |
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。 論文使用權限 Thesis access permission:校內校外完全公開 unrestricted 開放時間 Available: 校內 Campus: 已公開 available 校外 Off-campus: 已公開 available |
紙本論文 Printed copies |
紙本論文的公開資訊在102學年度以後相對較為完整。如果需要查詢101學年度以前的紙本論文公開資訊,請聯繫圖資處紙本論文服務櫃台。如有不便之處敬請見諒。 開放時間 available 已公開 available |
QR Code |