本論文主要研究以類神經網路模式預測季節性時間序列之有效性。利用適

當地建構樣本訓練集,網路經訓練後可作為季節性時間序列之預測工具。

文中亦提出移動學習法以期提高預測之準確度。並以台灣地區每季進口商

品與勞務總值則作為實證之研究。此季節性時間序列因受離群值之影響而

增加其預測困難度。實證結果顯示類神經網路模式之預測表現較傳統之統

計方法優異,即使此序列受到離群值之干擾。

We investigate the effectiveness of neural networks for

predicting the future behavior of seasonal time series.

Utilizing the training set constructed properly, we can train

the network who can be used to predict the future of seasonal

time series. A shifting-learning method is also employed in

order to obtained a better forecasting performance. The

quarterly imports of goods and services of Taiwan between the

first quarter of 1968 and the fourth quarter of 1990 are

studied in the research. The series are contaminated with

outliers, which will increase the difficulty of forecasting.

Empirical results exhibit that neural networks model free

approach have better prediction performance than the classical

Box-Jenkins approach, even the series are contaminated with

outliers.

Barron, A. R> (1991) Approximation bounds for superpositions of a sigmoid function. Proceedings of IEEE International Symposium on Information Theory. IEEE press.
Box, G. E. P. and Jenkins, G. G. (1976) Time Series Analysis : Forecasting and Control. 2nd-ed. San Francisco : Holden-Day.
Chen, C. And Liu, L. M. (1993) Forecasting time series with outliers. Journal of Forecasting, Vol. 12, 13-35.
Conradsen, K. and Spliid, H. (1981) A seasonal adjustment filter for use in Box Jenkins analyses of seasonal time sereis. Applied Statistics, Vol. 30, No.1, 172-177.
Cybenko, G. (1989) Approximation by superposition of a sigmoidal function. Methematics of control, signals and systems, Vol 2, 303-314.
Dagum, E. B. (1975) Seasonal factor forecasts from ARIMA models, paper presented at the 40th session of the International Statistic Institute, Warsaw, Poland.
Funahashi, K. I. (1989) On the approximate of continuous mappings by neural networks. Neural Networks, Vol. 2, 183-192.
Hecht-Nielsen, R. (1990) Neurocomputing. Addisom-Wesley.
Hornik, K., Stinchcombe, M. and White, H. (1989) Multilayer feedfoward networks are universal approximators, Neural Networks, Vol. 2, 359-366.
Lee, K. Y. , Cha, Y. T. and Part, J. H. (1991) Short-term load forecasting using an artificial neural network. IEEE Winter Power Meeting , paper No.91WM 199-0 PWRS.
Murthy, D. N. P. (1985) Parameter estimation with periodic observations. Journal of Time Sereis, Vol. 6, No.2. 91-96.
Osborn, D. (1991) The implications of periodically varying coefficients for seasonal time-series process. Journal of Econometrics, Vol. 48, 373-384.

Park, D. C. , El-Sharkawi, M. A. , Marks, R. J. , Atlas, L. E. and Damborg, M.J. (1990) Electric load forecasting using an artifical neural network. IEEE Summer Power Meeting, paper No.90SM 377-2 PWRS.
Rumelhart, D. E., Hinton, G. E. and Williams, R. J. (1986) Learning representations by back-propagation errors. Nature, Vol. 323, 533-536.
Vogl, T. P., Mangis, J. K., Rigler, A. K., Zink, W. T. and Alkon, D. L. (1988) Accelerating the convergence of the back-progation method. Biological Cybernetics, Vol.59, 257-263.
Weiglend, A., Huberman, B. And Rumelhar, D. (1990) Predicting the future : a connectionist approach. International Journal of Neural Systems, Vol. 1, No.3, 193-209.
Werbos, P. (1988) Generalization of back-propagation with application to recurrent gas market model. Neural Networks, Vol. 1. 339-256.
White, H. (1981) Consequences and detection of misspecified nonlinaer regulartion models. Journal of the American Statistic Association, Vol. 76, 419-433.
Wu, B. And Yu, C. (1992) Forecasting for nonstationary time series : a neural networks approach. Proceedings of the mid-America Chinese professional annual convention.