Decoupled echo state networks with lateral inhibition

In a real-life prediction task using noisy sea clutter data, both schemes exhibit higher prediction accuracy and successful design ratio than a conventional ESN with a single reservoir.

Abstract

Building on some prior work, in this paper we describe a novel structure termed the decoupled echo state network (DESN) involving the use of lateral inhibition. Two low-complexity implementation schemes, namely, the DESN with reservoir prediction (DESN + RP) and DESN with maximum available information (DESN + MaxInfo), are developed: (1) In the multiple superimposed oscillator (MSO) problem, DESN + MaxInfo exhibits three important attributes: lower generalization mean-square error (MSE), better robustness with respect to the random generation of reservoir weight matrix and feedback connections, and robustness to variations in the sparseness of reservoir weight matrix, compared to DESN + RP. (2) For a noiseless nonlinear prediction task, DESN + RP outperforms the DESN + MaxInfo and single reservoir-based ESN approach in terms of lower prediction MSE and better robustness to a change in the number of inputs and sparsity of the reservoir weight matrix. Finally, in a real-life prediction task using noisy sea clutter data, both schemes exhibit higher prediction accuracy and successful design ratio than a conventional ESN with a single reservoir.