Cluster-Based Input Weight Initialization for Echo State Networks

doi:10.1109/TNNLS.2022.3145565

Cluster-Based Input Weight Initialization for Echo State Networks

2023-10-01

10

OA

AI

DOI：10.1109/TNNLS.2022.3145565

Original

Original request for help

Save

Abstract

En 中文

Echo state networks (ESNs) are a special type of recurrent neural networks (RNNs), in which the input and recurrent connections are traditionally generated randomly, and only the output weights are trained. Despite the recent success of ESNs in various tasks of audio, image, and radar recognition, we postulate that a purely random initialization is not the ideal way of initializing ESNs. The aim of this work is to propose an unsupervised initialization of the input connections using the K-means algorithm on the training data. We show that for a large variety of datasets, this initialization performs equivalently or superior than a randomly initialized ESN while needing significantly less reservoir neurons. Furthermore, we discuss that this approach provides the opportunity to estimate a suitable size of the reservoir based on prior knowledge about the data.

Keywords:

Reservoirs

Neurons

Clustering algorithms

Task analysis

Self-organizing feature maps

Training

Mathematical models

Clustering

echo state networks (ESNs)

reservoir computing

unsupervised pretraining

Cluster-Based Input Weight Initialization for Echo State Networks

Abstract

Journal

Organization