臺灣博碩士論文加值系統

近年來，使用類神經網路進行匯率預測的文獻陸續被提出，然而這些文獻當中，並未全面性觀察何種類神經網路於預測匯率時是一較佳的網路模型，故本文嘗試使用十三種應用於倒傳遞網路之演算法，並且加上輻射基底函數網路與調適性網路模糊推論系統，這兩種不同型態的網路架構，以新台幣/美元匯率之每日收盤價格進行預測，希望藉此建構較佳的匯率預測模型。
研究結果顯示：一、輻射基底函數網路以及倒傳遞網路架構的隱藏層參數設定與平均絕對誤差百分比之間並無一規律性的改變，而調適性網路模糊推論系統的隸屬函數以使用高斯函數時，擁有較高的預測績效。二、輸入層神經元部分，各演算法與不同型態的網路架構，以使用一或二個輸入層神經元時，可獲得較佳預測績效。三、就訓練樣本數方面，各演算法與不同類型的類神經網路架構之間並不一致。四、整體的預測能力，以貝式規則化演算法預測績效較佳，其次依序為調適性網路模糊推論系統、輻射基底函數網路、Levenberg Marquardt演算法、BFGS擬牛頓法、比例共軛梯度法，以及一步階正割演算法。五、貝式規則化演算法配合使用222筆的訓練樣本、二個輸入層神經元以及十個隱藏層神經元的參數設定，可獲得較佳的預測結果，為本研究建議之類神經網路模型。

Recently, there are quite a few researches concentrate on forecasting exchange rate via neural network. Among these researches, a best neural network models in forecasting exchange rate is not found yet. So this research tries to use various Back-Propagation Network’s algorithms, Radial Basis Functions network and Adaptive Network-based Fuzzy Inference System to forecast prices of US/NT exchange rate with the expectation of presenting a better forecasting model.
The results are as follow; (1) In Back-Propagation Network, the number of hidden neurons doesn’t affect MAPE. As for Adaptive Network-based Fuzzy Inference System, Gauss Membership Function for the input variables, and the best parameters found. (2) Applying one or two input variables to get better performance. (3) Every neural network model which has different train samples. (4) Among these neural network, Bayesian Regularization Method performed best to forecast exchange rate, followed by Adaptive Network-based Fuzzy Inference System, Radial Basis Functions network, Levenberg-Marquardt Method, BFGS Quasi-Newton Method, Scaled Conjugate Gradient Method and One Step Secant Method in sequence. (5) Finally, Bayesian Regularization Method uses 222 train samples, two input variables and ten hidden neurons to obtain better performance.

誌謝iv
中文摘要v
英文摘要vi
目錄vii
圖目錄ix
表目錄x
第一章 緒論1
第一節 研究動機與目的1
第二節 本文架構5
第二章 文獻回顧6
第一節 類神經網路介紹6
第二節 倒傳遞類神經網路相關文獻7
第三節 其他型態之類神經網路相關文獻11
第三章 研究方法16
第一節 類神經網路模型介紹16
第二節 研究流程22
第三節 應用軟體與網路預測績效之評量標準24
第四章 實證結果25
第一節 樣本資料與參數設定25
第二節 類神經網路預測結果26
第三節 網路模型之比較與分析33
第五章 結論與建議 36
參考文獻38
附錄44
圖目錄
圖2.1 神經元模型6
圖3.1 倒傳遞網路架構16
圖3.2 輻射基底函數網路架構21
圖3.3 調適性網路模糊推論系統網路架構22
圖3.4 本研究流程圖23
圖4.1 樣本期間內新台幣/美元匯率趨勢圖25
圖4.2 輻射基底函數網路之匯率預測趨勢圖28
圖4.3 調適性網路模糊推論系統之匯率預測趨勢圖30
圖4.4 貝式規則演算法之匯率預測趨勢圖33
表目錄
表4.1 輻射基底函數網路於不同輸入層神經元個數之預測結果 27
表4.2 輻射基底函數網路於不同隱藏層中心點個數之預測結果 27
表4.3 調適性網路模糊推論系統於不同輸入層神經元個數之預測結果28
表4.4 調適性網路模糊推論系統於於不同隸屬函數類型之預測結果29
表4.6 倒傳遞網路演算法於不同輸入層神經元個數之預測結果 30
表4.7 倒傳遞網路演算法於不同隱藏層神經元個數之預測結果 32
表4.8 各種類神經網路於不同樣本數之預測結果34

1.中央銀行(2005)，中央銀行季刊，第二十七卷第四期。
2.張斐章、張麗秋與黄浩倫(2004)，類神經網路理論與實務，東華書局。
3.葉怡成(2001)，類神經網路模型應用與實作，儒林書局。
4.賴景昌(1993)，國際金融理論-基礎篇，茂昌書局，13~23頁。
5.李天行(2004)，「整合類神經網路與迴歸分析於匯率之預測－以東南亞金融風暴期間新台幣兌美元匯率為例」，統計與資訊評論， 1-24頁。
6.聶建中、馮正安與郭繼良(2001)，「金融風暴期間新台幣兌美元匯率預測－倒傳遞神經網路之應用」，台灣金融財務季刊，119~147頁。
7.卓師銘(2003)，遺傳演化類神經網路預測匯率─以美元為例，東吳大學經濟學研究所碩士論文。
8.徐希銘(2003)，運用類神經網路預測新台幣匯率，東吳大學企業管理研究所碩士論文。
9.邱志中(2003)，長短期匯率預測模式績效之比較，成功大學財務金融研究所碩士論文。
10.梁晉嘉(2002)，以非線性模式進行匯率走勢預測之研究－類神經網路模式之建立與應用，中山大學經濟學研究所碩士論文。
11.Battiti. R. (1992). “First- and Second-Order Methods for Learning between Steepest Descent and Newton’s Method.” Neural Computation, Vol. 4, no. 2, pp.141-166.
12.Bennie, W. and A. Milam (1999).“Pricing of Homeowner Association Dues with a Neural Network”, International Journal of Information and Management Sciences, Vol. 10, no.1, pp.73-80.
13.Broyden, G. C. (1970). “The Convergence of a Class of Double-Rank Minimization Algorithms”, Journal Institute of Mathematics. and Its Applications, Vol. 6, pp. 76-90.
14.Chinn, M. D. and R. A. Meese (1995). “Banking on Currency Forecasts: How Predictable is Change in Money? “, Journal of Internation Economics, Vol. 38, pp. 161-178.
15.Clarida, R. H., Sarno, L., Taylor, M. P. and G. Valente (2003). “The Out-of- Sample Success of Term Structure Models as Exchange Rate Predictors: A Step Beyond”, Journal of International Economics, Vol.60, pp. 61-83.
16.Coskun, N. and T. Yildirim (2003). “The Effects of Training Algorithms in MLP Network on Image Classification”, Neural Networks, 2003. Proceedings of the International Joint Conference, Vol.2, pp. 1223-1226.
17.David, J. T., Episcopos, A. and S. Wettimony (2001). “Predicting Direction Shifts on Canadian-US Exchange Rate with Artifical Neural Network”, International Journal of Intelligent Systems in Accounting, Vol.10, pp. 83-96.
18.Demuth, H. and M. Beale (2001). Neural Network Toolbox. For use with MATLAB. User’s guide, version 4.
19.Episcopos, A. and J. Davis (1996). “Predicting Returns on Canadian Exchange Rates With Artificial neural Networks and EGARCHM-M Model“, Neural Computing and Application, Vol. 4, pp. 168–174.
20.Fiore, C. D., Fanelli, S. and P. Zellini (2004). “An Efficient Generalization of Battiti-Shanno’s Quasi-Newton Algorithm for Learning in MLP-Networks”, Springer-Verlag Berlin Heidelberg, pp. 483-488.
21.Fine T. L. (1999). Feedforward Neural Network Methodology. Springer-Verlag.
22.Fletcher, R. and C. M. Reeves (1964). “Function Minimization by Conjugate Gradients”. Computer Journal 7, pp.149-154.
23.Fletcher, R. (1970). “A New Approach to Variable Metric Algorithms”, The Computer Journal, Vol. 13, pp. 317-322.
24.Goldfarb, D. (1970). “A Family of Variable Metric Methods Derived by Variational Means”, Mathematics Computation, Vol. 24, pp. 23-26.
25.Hann, T. H. and E. Steurer (1996). “Much ado About Nothing? Exchange Rate Forecasting: Neural Networks vs. Linear Models Using Monthly and Weekly Data”, Neurocomputing, Vol. 10, pp. 323-339.
26.Jang, J.-S. R. (1993). “ANFIS: Adaptive-Network-Based Fuzzy Inference Systems “ ,IEEE Transactions on Systems, Vol. 23, pp. 665-685.
27.Johan, F. K. and K. V. D. Herman (2002). “Neural Network Pruning Applied to Real Exchange Rate Analysis”, Journal of Forecasting,Vol. 21, pp. 559-577.
28.Kamrwzaman, J. and R. A. Sarke (2003). “Forecasting of Currency Exchange Rate Using ANN: A Case Study”, IEEE, December.14-17, pp. 793-797.
29.Levenberg, K. (1944). “A Method for the Solution of Certain Problems in Least Squares”, Quaterly of Applied Mathematics, pp.164-168.
30.Lewis, C. D. (1982). “Industrial and Business Forecasting Methods”, London, Butterworths.
31.Lin, C. T. and C. S. G. Lee (1996). Neural Fuzzy System. New Jersey：Prentice-Hall Inc.
32.Marquardt, D. (1963). "An Algorithm for Least-Squares Estimation of Nonlinear Parameters", Journal of the Society for Industrial and Apllied Mathematics, pp.431-441.
33.Makridakis, S. (1993) “Accuracy measures: Theoretical and practical concerns,” International Journal of Forecasting, Vol. 9, pp. 527-529.
34.MacKay, D. J. C. (1992) “A Practical Bayesian Framework for Backpropagation Networks”. Neural Computation, Vol. 4(3), pp. 448-472.
35.Moody, J. and C. Darken (1989). “Fast Learning in Networks of locally- tuned Processing Units”, Neural Computation, Vol. 1, pp. 281-294.
36.Moller, M. F. (1993). “A Scaled Conjugate Gradient Algorithm for Fast Supervised Learning”. Neural Network Vol. 6, pp. 525-533.
37.Nakagawa, H. (2002). “Real Exchange Rates and Real Interest Differentials: Implications of Nonlinear Adjustment in Real Exchange Rates“, Journal of Monetary Economics, Vol. 49, pp. 629-649.
38.Powell, M. J. D. (1977). “Restart Procedures for the Conjugate Gradient Method”. Mathematical Programming Vol. 12, pp.241-254.
39.Polak, E. and G. Ribiere (1969). “Note Sur La Convergence De Methodes De Directions Conjuguees”, Revue Francaise Informatique et de Recherche Operationelle. pp. 35-43.
40.Riedmiller, M. and H. Braun (1993). “A Direct Adaptive Method for Faster Backpropagation Learning:The RPROP Algorithm”, IEEE International Conference, pp.586-591.
41.Rumelhart, D. E., Hinton, G. E. and R. J. Williams (1986). “Learning Repres- entations by BackPropagation Errors”. Nature, Vol. 323, pp. 533-536
42.Staley, M. and p. Kim (1995). “Predicting The Canadian Spot Exchange Rate With Neural Networks”,Proceedings of the IEEE/IAFE Computational Intell- igence for Financial Engineering, pp.108-112.
43.Sfetsos, A. and C. Siriopoulos (2004). “Combinatorial Time Series Forecasting Based on Clustering Algorithms and Neural Nnetworks”, Neural Computing and Applications, Vol. 13, pp.56-64.
44.Shanno, D. F. (1970), “Conditioning of quasi-Neuwton methods for function minimization”, Mathematics Computation, Vol. 24, pp. 647-657.
45.Taylor, M. P. (1995). “The Economics of Exchange Rates”, Journal of Economic Literature, Vol. 33, pp. 13-47.
46.Wei, W. X. and Z. H. Jiang (1995). “Artificial Neural Network Forecasting Model for Exchange Rate and Empirical Analysis”, Forecasting, Vol. 2, pp.67–69.
47.Wei, H. (2004). “Forecasting Foreign Exchange Rates with Artificial Neural Networks: A Review”, International Journal of Information Technology & Decision Making, Vol. 3, No. 1, pp. 145–165
48.Weigend, A. S., Huberman, B. A. and D. E. Rumelhart (1992). “Predicting Sunspots and Exchange Rates with Connectionist Networks”, Nonlinear Modeling and Forecasting, pp. 395–432.
49.Wu, B. (1995). “Model-free Forecasting for Nonlinear Time Series (with Application to Exchange Rates) “, Computational Statistics and Data Analysis, Vol. 19, pp. 433–459.
50.Wu, I. F. and Y. J. Goo (2005). “A Neuro-Fuzzy Computing Technique for Modeling the Time Series of Short-Term NT$/US$ Exchange Rate “, The Journal of American Academy of Business, Vol. 7, Num 2, pp. 176-181.
51.Yao, J. T. and C. L. Tan (2000). “A Case Study on Using Neural Networks to Perform Technical Forecasting of Forex “, Neurocomputing, Vol. 34, pp. 79–98.
52.Zhang, G. and M. Y. Hu (1998). “Neural Network Forecasting of the British Pound /US Dollar Exchange Rate”, Journal of Management Science, Vol. 26, pp.495–506.
53.Zhang, G, Patuwo, B. E. and M. Y. Hu (1998). “Forecasting with Artificial Neural Networks: The State of the Art”, International Journal of Forecasting, Vol.14, pp.35–62.