This paper is dedicated to the long-term, or multi-step-ahead, time series prediction problem. We propose a novel method for training feed-forward neural networks, such as multilayer perceptrons, with tapped delay lines. Special batch calculation of derivatives called Forecasted Propagation Through Time and batch modification of the Extended Kalman Filter are introduced. Experiments were carried out on well-known timeseries benchmarks, the Mackey-Glass chaotic process and the Santa Fe Laser Data Series. Recurrent and feed-forward neural networks were evaluated. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Chernodub, A. (2013). Direct method for training feed-forward neural networks using batch extended kalman filter for multi-step-ahead predictions. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8131 LNCS, pp. 138–145). https://doi.org/10.1007/978-3-642-40728-4_18
Mendeley helps you to discover research relevant for your work.