A State Space Approach and Hurst Exponent for Ensemble Predictors
In this article we propose a concept of ensemble methods based on deconvolution with state space and MLP neural network approach. Having a few prediction models we treat their results as a multivariate variable with latent components having destructive or constructive impact on prediction. The latent component classification is performed using novel variability measure derived from Hurst exponent. The validity of our concept is presented on the real problem of load forecasting in the Polish power system.
Keywordsstate space approach Hurst exponent Independent Component Analysis ensemble methods
Unable to display preview. Download preview PDF.
- 1.Barabasi, A.-L., Vicsek, T.: Multifractality of self-affine fractals. Physical Review A44, 2730–2733 (1991)Google Scholar
- 3.Bussgang, J.J.: Cross-correlation function of amplitude-distorted Gaussian signals, MIT Research Laboratory of Electronics, Technical Report 216 (1952)Google Scholar
- 7.Hurst, H.E.: Long term storage capacity of reservoirs. Trans. Am. Soc. Civil Engineers 116, 770–799 (1951)Google Scholar
- 8.Peters, E.: Fractal market analysis. John Wiley, Chichester (1996)Google Scholar
- 9.Samorodnitskij, G., Taqqu, M.: Stable non-Gaussian random processes: stochastic models with infinitive variance. Chapman and Hall, New York (1994)Google Scholar
- 13.Zhang, L., Cichocki, A.: Blind Separation of Filtered Source Using State-Space Approach. In: Advances in Neural Information Processing Systems, vol. 11, pp. 648–654 (1999)Google Scholar
- 14.Zhang, L., Cichocki, A.: Blind Deconvolution of Dynamical Systems: A State Space Approach. Journal of Signal Processing 4(2), 111–113 (2000)Google Scholar