本研究使用複數模糊類神經系統 (Complex neuro-fuzzy system, CNFS) 結合支持向量迴歸 (Support vector regression, SVR) 為方法，設計一個新穎的複數模糊類神經支持向量迴歸預測器 (Complex neuro-fuzzy system based SVR, CNFS-SVR)，並結合移動區塊拔靴法作時間序列的區間預測之研究。複數模糊類神經系統使用複數模糊集合，複數模糊集合擁有優秀的適應能力與非線性映射能力，且複數模糊類神經系統的輸出是屬於複數值，預測值的實部與虛部可以用於同時預測兩個不同的時間序列。支持向量迴歸理論基於結構風險最小化 (Structural risk minimization, SRM) 原則設計以達到優秀的一般化能力。模糊類神經系統作為支持向量迴歸器的映射函數，如此複數模糊類神經支持向量迴歸預測器經由模糊類神經系統與支持向量迴歸器的協同作用來達到系統優異的預測性能。在系統學習階段中，模糊C平均分裂演算法可以自動產生適當的規則數作為系統架構，並使用粒子群最佳化演算法與遞迴最小平方估計法，形成複合式學習法進行系統參數學習。根據移動區塊拔靴法與該預測器來建立信賴區間對時間序列作區間預測。本研究使用各種不同的匯率作為研究對象，其結果顯示複數模糊類神經支持向量迴歸預測器結合移動區塊拔靴法能同時預測兩個時間序列且期預測性能有不錯的表現。

A novel intelligent approach using complex neuro-fuzzy system based support vector regression (denoted as CNFS-SVR) and moving-block bootstrap is proposed to the problem of time series interval forecasting in this thesis. The proposed CNFS-SVR approach combines both of the complex neuro-fuzzy system (CNFS) theory and the support vector regression (SVR) theory. With complex fuzzy sets (CFSs), the CNFS has excellent adaptive ability for functional mapping. The output of CNFS is complex-valued and can be used to develop the so-called dual output capability, which can be used to predict two time series simultaneously. SVR is based on the statistical learning theory. With the principle of structural risk minimization (SRM), SVR can possess excellent generalization ability without over fitting. In the study, CNFS-SVR is developed to integrate the merits of the CNFS theory and the SVR rationale to obtain excellent performance. Bootstrap is a re-sampling method, by which empirical statistical distribution can be developed and confidence interval can be obtained using statistical inference. For the learning strategy, a FCM-based clustering method is used to automatically determine the initial knowledge base of CNFS-SVR. Particle swarm optimization (PSO) and recursive least squares estimator (RLSE) algorithm are used in a hybrid way to update the parameters of If-Then fuzzy rules of CNFS-SVR. The LibSVM package is used to optimize the proposed CNFS-SVR machines. Several real-world exchange-rate time series are used in the study. The experimental results show promising performance.

[1] 王黎明、王連和楊楠，應用時間序列分析，初版，復旦大學出版社，上海，2009年。
[2] 王燕，應用時間序列分析，初版，中國人民大學出版社，北京，2005年。
[3] Madsen, H., Time Series Analysis, 1st edition, Chapman and Hall/CRC, 2008.
[4] Taylor, S.J., Modelling Financial Time Series, 2nd edition, World Scientific Publishing Company, 2007.
[5] Janacek G., Time Series Analysis: Forecasting and Control, 4th edition, John Wiley and Sons, Inc., 1994.
[6] Wagner, N., Michalewicz, Z., Khouja, M. and McGregor, R.R., “Time series forecasting for dynamic environments: the DyFor genetic program model,” IEEE Transactions on Evolutionary Computation, vol. 11, issue 4, pp. 433-452, 2007.
[7] Esfahanipour, A. and Aghamiri, W., “Adapted neuro-fuzzy inference system on indirect approach TSK fuzzy rule base for stock market analysis,” Expert Systems with Applications, vol. 37, pp. 4742-4748, 2010.
[8] Chen, Y., Yang, B., Dong, J. and Abraham, A., “Time-series forecasting using flexible neural tree model,” Information Sciences, vol. 174, pp. 219-235, 2005.
[9] Cho, K.B. and Wang, B.H., “Radial basis function based adaptive fuzzy systems and their applications to system identification and prediction,” Fuzzy Sets and Systems, vol. 83, pp. 325-339, 1996.
[10] Jilani, T.A., Burney, S.M.A. and Ardil, C., “Fuzzy metric approach for fuzzy time series forecasting based on frequency density based partitioning,” International Journal of Computational Intelligence, vol. 4, pp. 112-117, 2007.
[11] Sugiarto, I. and Natarajan, S., “Parameter estimation using least square method for MIMO Takagi-Sugeno neuro-fuzzy in time series forecasting,” Jurnal Teknik Elektro, vol. 7, no. 2, 2008.
[12] Molodtsova, T. and Papell, D.H., “Out-of-sample exchange rate predictability with Taylor rule fundamentals,” Journal of International Economics, vol. 77, pp. 167-180, 2009.
[13] Leu, Y., Lee, C.P. and Jou, Y.Z., “A distance-based fuzzy time series model for exchange rates forecasting,” Expert Systems with Applications, vol. 36, pp. 8107-8114, 2009.
[14] Goldberg, E., “Optimal signal multi-resolution by genetic algorithms to support artificial neural networks for exchange-rate forecasting,” Journal of Political Economy, vol. 104, pp. 510-541, 1996.
[15] Zhang, G. and Hu, M.Y., “Neural network forecasting of the British pound/US dollar exchange rate”, Omega, vol. 26, pp. 495-506, 1998.
[16] Luna, I., Soares S. and Ballini, R., “A constructive-fuzzy system modeling for time series forecasting,” IEEE International Joint Conference on Neural Networks, vol. 16, pp. 2908-2913, 2007.
[17] Zhao, L. and Yang, Y.P., “PSO-Based single multiplicative neuron model for time series prediction,” Expert Systems with Applications, vol. 36, pp. 2805-2912, 2009.
[18] Rojas, I., Valenzuela, O., Rojas F., Guillen, A., Herrera, L.J. Pomares, H., Marquez, L. and Pasadas, M., “Soft-computing techniques and ARMA model for time series prediction,” Neurocomputing, vol. 71, pp. 519-537, 2008.
[19] Jang, S.R., “ANFIS: adaptive-network-based fuzzy inference system,” IEEE Transactions on System, Man, and Cybernetics, vol. 23, pp. 665-685, 1993.
[20] Sfetsos A. and Siriopoulos C., “Time series forecasting with a hybrid clustering scheme and pattern recognition,” IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans, vol. 34, pp. 399-405, 2004.
[21] Herrera, L.J., Pomares, H., Rojas, I., Guillén, A., González, J., Awad, M. and Herrera, A., “Multigrid-based fuzzy systems for time series prediction: CATS competition,” Neurocomputing, vol. 70, issue 13-15, pp. 2410-2425, 2007.
[22] Thawonmas, R. and Abe, S., “Function approximation based on fuzzy rules extracted from partitioned numerical data,” IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, vol. 29, pp. 525-534, 1999.
[23] Hornik, K., “Multilayer feedforward networks are universal approximators,” Neural Networks, vol. 2, issue 5, pp. 359-366, 1989.
[24] Kandel, A., Ramot, D., Milo, R. and Friedman, M., “Complex fuzzy sets,” IEEE Transactions on Fuzzy Systems, vol. 10, pp. 171–186, 2002.
[25] Dick, S., “Toward complex fuzzy logic,” IEEE Transactions on Fuzzy Systems, vol. 13, pp. 405-414, 2005.
[26] Ramot, D., Friedman, M., Langholz, G. and Kandel, A., “Complex fuzzy logic,” IEEE Transactions on Fuzzy Systems, vol. 11, pp. 450-461, 2003.
[27] Li, C. and Chiang, T.W., “Complex fuzzy computing to time series prediction – A multi-swarm PSO learning approach,” Lecture Notes in Artificial Intelligence, vol. 6592, pp. 242-251, 2011.
[28] Li, C., Chiang T.W., Hu J.W. and Wu T., “Complex neuro-fuzzy intelligent approach to function approximation,” in proceedings: 2010 Third International Workshop on Advanced Computational Intelligence (IWACI), pp. 151-156, 2010.
[29] Chen, Y., Yang, B. and Dong, J., “Time-series prediction using a local linear wavelet neural network,” Neurocomputing, vol. 69, pp. 449-465, 2006.
[30] Wong, W.K., Xia, M. and Chu, W.C., “Adaptive neural network model for time-series forecasting,” European Journal of Operational Research, vol. 207, pp. 807-816, 2010.
[31] Smola, A. and Schölkopf, B., “A tutorial on support vector regression,” NeuroCOLT Technical Report NC-TR-98-030, Royal Holloway College, University of London, UK, 1998.
[32] Burges, C.J.C., “A tutorial on support vector machines for pattern recognition,” Data Mining and Knowledge Discovery, vol. 2, no. 2, pp. 121-167, 1998.
[33] Jeng, J.T., Chuang, C.C. and Tao, C.W., “Hybrid SVMR-GPR for modeling of chaotic time series systems with noise and outliers,” Neurocomputing, vol. 73, pp. 1686-1693, 2010.
[34] Wang, L. and Mu, Z.C., “A new support vector neural network inference system,” in proceedings: International Federation for Information Processing, 2005(IFIP 2005), vol. 163, pp. 485-494, 2005.
[35] Christoffersen, P.F., “Evaluating interval forecasts,” International Economic Review, vol. 39, no. 4, pp. 841-862, 1998.
[36] Kim, J.H., Wong, K., Athanasopoulos, G. and Liu S., “Beyond point forecasting: Evaluation of alternative prediction intervals for tourist arrivals,” International Journal of Forecasting, vol. 27, issue 3, pp. 887-901, 2010.
[37] Chatfield, C., “Calculating interval forecasts,” Journal of Business and Economic Statistics, vol. 11, no. 2, pp. 121-135, 1993.
[38] Clements, M.P. and Taylor, N., “Evaluating interval forecasts of high-frequency financial data,” Journal of Applied Econometrics, vol. 18, issue 4, pp. 445-456, 2003.
[39] Tsaur, R.C., “The development of an interval grey regression model for limited time series forecasting,” Expert Systems with Applications, vol. 37, issue 2, pp. 1200-1206, 2010.
[40] Maia, A.L.S. and Carvalho, F.A.T., “Holt''s exponential smoothing and neural network models for forecasting interval-valued time series,” International Journal of Forecasting, vol. 27, issue 3, pp. 740-759, 2010.
[41] Liu, R.Y. and Singh, K., Exploring the Limits of Bootstrap, John Wiley and Sons, 1st edition, Wiley-Interscience, pp. 225-248, 1992.
[42] Efron, B. and Tibshirani, R.J., “Bootstrap methods for standard errors, confidence intervals, and other measures of statistical accuracy,” Statistical Science, vol. 1, pp. 54-75, 1986.
[43] Politis, D.N., “The impact of bootstrap methods on time series analysis,” Statistical Science, vol. 18, no. 2, pp. 219-230, 2003.
[44] Jang, J.S.R., Sun, C.T. and Mizutani, E., Neuro-Fuzzy and Soft Computing: a Computational Approach to Learning and Machine Intelligence, USA: Prentice Hall, 1997.
[45] Sun, H., Wang, S. and Jiang, Q., “FCM-based model selection algorithms for determining the number of clusters,” Pattern Recognition, vol. 37, pp. 2027-2037, 2004.
[46] Kennedy, J. and Eberhart, R., “Particle swarm optimization,” in proceedings: IEEE International Conference on Neuro Network Proceedings, vol. 4, pp. 1942-1948, 1995.
[47] Astrom K.J. and Wittenmark, B., Adaptive Control, Addison-Wesley Longman Publishing Co., Inc. Boston, MA, USA, 1994.
[48] Vapnik, V., Golowich, S.E. and Smola, A., “Support vector method for function approximation, regression estimation, and signal processing,” in proceedings: Advances in Neural Information Processing Systems 9, 281-287, 1997.
[49] Gunn, S.R., “Support vector machines for classification and regression,” Technical Report, Image Speech and Intelligent Systems Research Group, University of Southampton, UK., 1997.
[50] Pal, B.D. and Patranabis D.C., “Support vector regression,” Neural Information Processing—Letters and Reviews, vol. 11, no. 10, pp. 203-224, 2007.
[51] Drucker, H., Burges, C.J.C., Kaufman, L, Smola, Alex and Vapnik V., “Support vector regression machines,” Advances in Neural Information Processing Systems, Vol. 9, pp.155-161, 1997.
[52] Zadeh, L.A., “Fuzzy sets,” Information and Control, vol. 8, pp. 338-353, 1965.
[53] Zadeh, L.A., “Fuzzy logic and approximate reasoning,” Synthese, vol. 30, pp. 407-428, 1975.
[54] Buckley, J.J., “Fuzzy complex numbers,” Fuzzy Sets and Systems, vol. 33, pp. 333-345, 1989.
[55] Buckley, J.J. and Qu, Y., “Fuzzy complex analysis I: Differentiation,” Fuzzy Sets and Systems, vol. 41, pp. 269-284, 1991.
[56] Buckley, J.J., “Fuzzy complex analysis II: Integration,” Fuzzy Sets and Systems, vol. 49, pp. 171-179, 1992.
[57] Horowitz J.L., “The bootstrap,” Handbook of Econometrics, vol. 5, pp. 3159-3228, 2001.
[58] Carlstein, E., Do, K.A., Hall, P., Hesterberg, T. and Künsch, H.R. “Matched-block bootstrap for dependent data,” Bernoulli, vol. 4, no.3, pp. 305-328, 1998.
[59] Vogel, R.M. and Shallcross, A.L., “The moving blocks bootstrap versus parametric time series models,” Water Resources Research, vol. 32, no. 6, pp. 1875-1882, 1996.
[60] Dunn, J.C., “A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters,” Cybernetics and Systems, vol. 3, pp. 32-57, 1973.
[61] Ramze, Rezaee, M., Lelieveldt, B.P.F. and Reiber, J.H.C., “A new cluster validity index for the fuzzy c-mean,” Pattern Recognition Letters, vol. 19, pp. 237-246, 1997.
[62] Holland, J.H., Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence, University of Michigan Press, Ann Arbor, 1975.
[63] Kirkpatrick, S., Gelatt, C.D. and Vecchi, M.P., “Optimization by simulated annealing,” Science, vol. 220, pp. 671-680, 1983.
[64] Wang, L.X. and Mendel, J.M., “Generating fuzzy rules by learning from examples,” IEEE Transactions on Systems, Man and Cybernetics, vol. 22, no. 6, 1414-1427, 1992.
[65] Chen, Y. and Wang, J.Z., “Kernel machines and additive fuzzy systems: classification and function approximation,” in proceedings: 12th IEEE International Conference on Fuzzy Systems, vol. 2, pp. 789-795, 2003.
[66] Lin, C.T., Liang, S.F., Yeh, C.M. and Fan, K.W., “Fuzzy neural network design using support vector regression for function approximation with outliers,” in proceedings: IEEE International Conference on System, Man and Cybernetics, vol. 3, pp. 2763-2768, 2005.
[67] Juang C.F. and Hsieh C.D., “TS-fuzzy system-based support vector regression,” Fuzzy Sets and Systems, vol. 160, issue 17, pp. 2486-2504, 2009.
[68] Bordo, M.D. and Eichengreen, B.J., A Retrospective on the Bretton Woods System: Lessons for International Monetary Reform, University of Chicago Press, USA, 1993.
[69] He, L.T. and Hu, C., “Impacts of interval computing on stock market variability forecasting,” Computational Economics, vol. 33, number 3, pp. 263-276, 2009.
[70] Chang, C.C. and Lin, C.J., LIBSVM: A library for support vector machines. [Online] Available: (http://www.csie.ntu.edu.tw/~cjlin/libsvm/).
[71] OANDA, http://www.oanda.com/currency/historical-rates