Evolutionary Multi-Model Estimators for ARMA System Modeling and Time Series Prediction

  • Authors:

  • Grigorios Beligiannis;Spiridon Likothanassis;Lambros Skarlas

  • Affiliations:

  • Department of Computer Engineering & Informatics, University of Patras, Patras, Greece GR-26500 and Computer Technology Institute (C.T.I.), Patras, Greece GR-26221;Department of Computer Engineering & Informatics, University of Patras, Patras, Greece GR-26500 and Computer Technology Institute (C.T.I.), Patras, Greece GR-26221;Department of Computer Engineering & Informatics, University of Patras, Patras, Greece GR-26500 and Computer Technology Institute (C.T.I.), Patras, Greece GR-26221

  • Venue:
  • IWANN '03 Proceedings of the 7th International Work-Conference on Artificial and Natural Neural Networks: Part II: Artificial Neural Nets Problem Solving Methods
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this work, a well-tested evolutionary method for system modeling and time series prediction is presented. The method combines the effectiveness of adaptive multi model partitioning filters and GAs' robustness. Specifically, the a posteriori probability that a specific model, of a bank of the conditional models, is the true model can be used as fitness function for the GA. In this way, the algorithm identifies the true model even in the case where it is not included in the filters' bank and is able to accurately forecast the shortterm evolution of the system. The method is not restricted to the Gaussian case; it is computationally efficient and is applicable to on-line/adaptive system modeling and time series prediction.