以遺傳演算法優化支援向量迴歸在旅遊需求預測上的應用

Using Genetic Algorithms to Optimize Support Vector Regression in Tourism Demand Forecasting

10.6130/JORS.2005.18(1)3

陳寬裕(Kuan-Yu Chen)；王正華(Cheng-Hua Wang)

支援向量機 ； 類神經網路 ； 旅遊需求 ； 敏感度分析 ； 過度擬合 ； Support vector machine ； artificial neural networks ； tourism demand ； sensitivity analysis ； over-fitting

戶外遊憩研究

18卷1期（2005 / 03 / 01）

47 - 72

繁體中文

支援向量機（support vector machine，簡稱SVM）的發展最先被運用在模式識別領域，然而隨著ε-不敏感損失函數(ε-insensitive loss function)的導入，支援向量機已經被擴展到解決非線性迴歸估計的問題上，此類技術稱爲支援向量迥歸（support vector regression，簡稱SVR）。本研究將運用支援向量迴歸技術建構旅遊需求之預測模型。研究中將提出一種名爲GA-SVR的新模型，該模型先運用實數值遺傳演算法以找出支援向量迥歸的最佳化參數，並籍這些最佳化參數值建構支援向量迥歸模型，以預測旅遊需求，並將其預測結果和類神經網路模型(artificial neural networks)進行比較，以證明GA-SVR模型確實擁有優良的預測能力。此外爲了驗證參數對支援向量迥歸模型的重要性並獲知支援向量迴歸模型的一些特性，研究中使用了敏感度分析技術(sensitivity analysis)，該分析中也證實了，不當的選取參數將使模型容易陷矜過度擬合(over-fitting)或不足擬合(under-fitting)的危機中。

Support Vector Machine (SVM) was first applied to pattern recognition problem, however, with the introduction of £'-insensitive loss function by Vapnik, SVM has been applied to non-linear regression estimation, a new technique called Support Vector Regression (SVR). This study will apply support vector regression techniques to construct the predictive model of tourism demand. In order to construct an effective SVR model, we had to set SVR's parameters carefully. This study proposes a new model called GA-SVR, that is searching for SVR's optimal parameters through applied real-valued genetic algorithms, and which uses optimal parameters to construct a SVR model. According to the results of the tourism demand forecasting study, and SVR model shows reliability and good prediction. Its generalized performance is even more accurate than artificial neural networks. Moreover, in order to test the importance and understand the features of the SVR model, this study implemented a sensitivity analysis technique. The analysis demonstrates that improperly selected parameters will cause the over-fitting or under-fitting problem in the SVR model.

1. Campbell, C.(2002).Demel methods: A survey of current techniques.Neurocomputing.
2. Chan, Y. M.(1993).Forecasting tourism: A sine time series regression approach.Journal of Travel Research.
3. LIBSJM:A library for support vector machines
4. Cheng, W.,Wanger, L.,Lin, C. H.(1996).Forecasting the 30-year US treasury bond with a system of neural networks.Journal of Computational Intelligence in Finance.
5. Chiang W. C.,Urban, T. L.,Baildridge, G.(1996).A neural network approach to mutual fund net asset value forecasting.Omega.
6. Chtioui,Y.,Bertrand, D.,Barba, D.(1998).Feature selection by a genetic algorithm application to seed discrimination by artificial vision..Journal of Sci. Food Agric.
7. Chu, Fong-Lin.(1998).Forecasting tourism: A combined approach.Tourism Management,19(6),515-520.
8. Cristianini, N.,Campell, C.,Shawe-Taylor, J.(1999).Dynamically adapting kernels in support vector machines.Advances in neural information processing systems.
9. Cybenko, G.(1989).Approximation by superpositions of a sigmoidal function.Mathematical Control Signal System.
10. Deb, K,Agrawal, R. B.(1995).Simulated binary crossover for continuous search space.Complex Systems.
11. Deb, K,Goyal M.(1996).A combined genetic adaptive search (geneAS) for engineering design.Computer Science and Informatics.
12. Deb, K,Kumar, A.(1995).Real-coded genetic algorithms with simulated binary crossover: Studies on multimodal and multiobjective problems.Complex Systems.
13. Duan, K.,Keerthi, S.,Pun, A.(2001).Evaluation of simple performance measures for tuning SVM hyperparameters(Technical Report).
14. Goldberg, G.E.(1989).Genetic algorithms in search, optimization and machine learning.
15. Haupt, R. L.,Haupt, S. E.(1998).Practical genetic algorithms.
16. Haykin, S.(1999).Neural networks: A comprehensive foundation.
17. Holland, J. H.(1975).Adaptation in natural and artificial systems.
18. Homik, K.,Stinchcombe, M.,White, H.(1989).Multilayer feedforward networks are universal approximations.Neural Networks.
19. Text categorization with support vector machines: Learning with many relevant features (Technical Report)
20. Kaastra, I.,Boyd, M. S.(1995).Forecasting futures trading volume using neural networks.The Journal of Futures Markets,15(8),853-970.
21. Kulendran, N.,King, Maxwell.(1997).Forecasting international quarterly tourist flows using error-correction and time-series model.International Journal of Forecasting.
22. Law, R.(2000).Back-propagation learning in improving the accuracy of neural network-based tourism demand forecasting.Tourism Management,21(4),331-340.
23. Law, R.,Au, N.(1999).A neural network model to forécast Japanese demand for travel to Hong Kong.Tourism Management,20(1),89-97.
24. Lewis, C.D.(1982).International and business forecasting methods.
25. Lim, C.(1997).An econometric classification and review of international tourism demand models.Tourism Economics,3(1),69-81.
26. Manuel, V.,Croes, R.(2000).Evaluation of demand US tourists to Aruba.Annals of Tourism Research.
27. Martin, Christine A.,Witt, S.F.(1989).Forecasting tourism demand: A comparison of the accuracy several quantitative methods.International Journal of Forecasting.
28. Michalewicz, Z.(1999).Genetic algorithms+Data structures.
29. Pattie, Douglas C.,Snyder, John.(1996).Using a neural network to forecast visitor behavior.Annals of Tourism Research.
30. Pilar, Gonzalez.,Moral, Paz.(1995).An analysis of the international tourism demand in Spain.International Journal of Forecasting.
31. Rumelhart, E.,Hinton G. E.,Williams, R. J.(1986).Learning internal representations by error propagation in parallel distributed processing.
32. Schmidt, M.(1996).Identifying speaker with support vector networks.
33. Shards, P.,Patil, R.B.(1992).Connectionist approach to time series prediction: An empirical test.Journal of Intelligent Manufacturing.
34. Sheldon, Pauline J.,Var, Turgut.(1985).Tourism forecasting: A review of empirical research.Journal of Forecasting.
35. Smeral, K, Witt, S.F.,Witt, C. A.(1992).Econometric forecasts: Tourism trends to 2000.Annals of Tourism Research.
36. Smola AJ.(1998).Learning with kernels.
37. Tang K.S.,Man, K.F.,Kwong, S.,He, Q.(1996).Genetic algorithms and their applications.IEEE Signal Processing Magazine.
38. Tay, Francis. E. H.,Cao, Lijuan.(2001).Application of support vector machines in financial time series forecasting.Omega.
39. Uysal, M.,Crompton, J.L.(1984).Determinants of demand for international tourist flow to Turkey.Tourism Management,15(4),288-297.
40. Van, B.,Robert, J.(1997).The application of neural networks in the forecasting of share prices.Haymarket; VA:Finance & Technology Publishing.
41. Vapnik, V.(1995).The nature of statistical learning theory.New York:Springer.
42. Vapnik, V.,Golowich, S.,Smola, A.(1997).Support vector method for function approximation, regression estimation and signal processing.Advance in Neural information processing system.
43. 江麗文(1995)。台北，中國文化大學觀光事業研究所。
44. 李旭煌(1994)。台北，政治大學統計研究所。
45. 林保村(2001)。台北，台灣科技大學電機工程研究所。
46. 時巧偉(1994)。台北，政治大學統計研究所。
47. 曹勝雄、江麗文、蔡倩雯(1996)。計量經濟模式在國際觀光需求應用之研究。戶外遊憩研究，9(1)，1-29。
48. 曹勝雄、曾國雄、江勁毅(1996)。傳統計量迴歸、GMDH、類神經網路四種方法在預測應用之比較－以國人赴港旅客需求之預測爲例。中國統計學報，34(2)，132-161。
49. 陳石麟(2003)。台北，台灣大學資訊管理研究所。

1. 顏上堯,林政志,王俊懿(2020)。石門水庫服務台及售票站人力排班最佳化研究。台灣管理學刊，20(1)，99-127。