In this paper, we formally deduce a new computational model, with a recurrent structure, by means of data granulation. The proposed scheme can be regarded as an Echo State Network (ESN), with an additional granular layer. ESNs have been recently revisited in the context of deep learning. In view of such a state-of-the-art, and coherently with the concept of data granulation, the aim herein is to propose a more efficient and transparent structure. The stability of the proposed scheme is formally discussed. The performance is shown by means of several benchmarks against the state-of-the-art methods. The proposed architecture exhibits a lower computational cost and a higher accuracy.
Revising recurrent neural networks from a granular perspective
Colace, Francesco;Loia, Vincenzo
;Tomasiello, Stefania
2019
Abstract
In this paper, we formally deduce a new computational model, with a recurrent structure, by means of data granulation. The proposed scheme can be regarded as an Echo State Network (ESN), with an additional granular layer. ESNs have been recently revisited in the context of deep learning. In view of such a state-of-the-art, and coherently with the concept of data granulation, the aim herein is to propose a more efficient and transparent structure. The stability of the proposed scheme is formally discussed. The performance is shown by means of several benchmarks against the state-of-the-art methods. The proposed architecture exhibits a lower computational cost and a higher accuracy.File | Dimensione | Formato | |
---|---|---|---|
Colace Francesco 1-138 DEFINITIVO.pdf
non disponibili
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
1.35 MB
Formato
Adobe PDF
|
1.35 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
138 Colace Post-Print.pdf
accesso aperto
Descrizione: 1568-4946/© 2019 Elsevier B.V. All rights reserved. Link Editore: https://doi.org/10.1016/j.asoc.2019.105535
Tipologia:
Documento in Post-print (versione successiva alla peer review e accettata per la pubblicazione)
Licenza:
Creative commons
Dimensione
1.37 MB
Formato
Adobe PDF
|
1.37 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.