Cluster-Based Input Weight Initialization for Echo State Networks
IEEE Transactions on Neural Networks and Learning Systems · 2023-10 · 34 (10) · 7648-7659
DOI: 10.1109/tnnls.2022.3145565 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Neural Networks and Learning Systems, 2023-10. Verfasst von Peter Steiner und 2 weiteren Autor:innen. Die Arbeit wurde bisher 18-mal zitiert (Crossref) und nennt 42 Literaturangaben. Gefördert von: Europäischer Sozialfonds, Freistaat Sachsen, Special Research Fund of Ghent University.
Abstract (Originalsprache)
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.
Förderung
- Europäischer Sozialfonds
- Freistaat Sachsen
- Special Research Fund of Ghent University
Zitieren
APA
Steiner, P., Jalalvand, A., & Birkholz, P. (2023). Cluster-Based Input Weight Initialization for Echo State Networks. IEEE Transactions on Neural Networks and Learning Systems, 34(10), 7648-7659. https://doi.org/10.1109/tnnls.2022.3145565
Vancouver
Steiner P, Jalalvand A, Birkholz P. Cluster-Based Input Weight Initialization for Echo State Networks. IEEE Transactions on Neural Networks and Learning Systems. 2023;34(10):7648-7659. doi:10.1109/tnnls.2022.3145565
BibTeX
@article{Steiner2023,
author = {Steiner, Peter and Jalalvand, Azarakhsh and Birkholz, Peter},
title = {Cluster-Based Input Weight Initialization for Echo State Networks},
journal = {IEEE Transactions on Neural Networks and Learning Systems},
year = {2023},
volume = {34},
number = {10},
pages = {7648--7659},
doi = {10.1109/tnnls.2022.3145565}
}
EUVIMED