Anti boundary effect wavelet decomposition echo state networks

Abstract

We propose a novel approach based on wavelet decomposition and echo state networks to discover the multiscale dynamics of time series which we call anti-boundary-effect wavelet decomposition and echo state networks (ABE-WESNs). ABE-WESNs use the wavelet decomposition as preprocessing steps and choose a matched ESNs for every scale level. We use the data extension methods to overcome the boundary effect. The introduced weight factors can both resolve the problem of cumulation of errors resulting from the wavelet decomposition. Experiments and engineering applications show that the ABE-WESNs can accurately model and predict some time series with multiscale properties. © 2011 Springer-Verlag.