Strategies for Sequential Prediction of Stationary Time Series
We present simple procedures for the prediction of a real valued sequence. The algorithms are based on a combination of several simple predictors. We show that if the sequence is a realization of a bounded stationary and ergodic random process then the average of squared errors converges, almost surely, to that of the optimum, given by the Bayes predictor. We offer an analog result for the prediction of stationary gaussian processes.
KeywordsGaussian Process Modeling Uncertainty Ergodic Theorem Prediction Strategy Stationary Time Series
Unable to display preview. Download preview PDF.
- Bailey, D. H. (1976). Sequential schemes for classifying and predicting ergodic processes. PhD thesis, Stanford University.Google Scholar
- Gerencsér, L. and J. Rissanen. (1986). A prediction bound for Gaussian ARMA processes. Proc. of the 25th Conference on Decision and Control, 1487–1490.Google Scholar
- Goldenshluger, A. and A. Zeevi. (submitted for publication 1999). Non-asymptotic bounds for autoregressive time series modeling.Google Scholar
- Nobel, A. (2000). Aggregate schemes for sequential prediction of ergodic processes. Manuscript.Google Scholar
- Singer, A. C. and M. Feder. (2000). Universal linear least-squares prediction. International Symposium of Information Theory.Google Scholar
- Vovk, V.G. (1990). Aggregating strategies. In Proceedings of the Third Annual Workshop on Computational Learning Theory, pages 372–383. Association of Computing Machinery, New York.Google Scholar
- Yang, Y. (1999). Aggregating regression procedures for a better performance. Manuscript.Google Scholar