Recurrent Neural Networks as Local Models for Time Series Prediction

  • Authors:
  • Aymen Cherif;Hubert Cardot;Romuald Boné

  • Affiliations:
  • Laboratoire d'informatique, Université François Rabelais Tours, Tours, France 37200;Laboratoire d'informatique, Université François Rabelais Tours, Tours, France 37200;Laboratoire d'informatique, Université François Rabelais Tours, Tours, France 37200 and Ecole Nationale d'Ingénieurs du Val de Loire, Blois cedex, France 41034

  • Venue:
  • ICONIP '09 Proceedings of the 16th International Conference on Neural Information Processing: Part II
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Local Models for regression have focused a great deal of attention in recent years. They have proved to be more efficient than global models and especially when dealing with chaotic time series. Many models have been proposed to cluster time series and they have been combined with several predictors. In this paper we present an extension for recurrent neural networks in allowing to apply them to local models and we discuss the obtained results.