Thesis etd-03202025-113915 |
Link copiato negli appunti
Thesis type
Tesi di laurea magistrale
URN
etd-03202025-113915
Thesis title
Deep Reservoir Memory Network
Department
INFORMATICA
Course of study
INFORMATICA
Supervisors
.
relatore Prof. Gallicchio, Claudio
correlatore Prof. Ceni, Andrea
controrelatore Prof. Cossu, Andrea
correlatore Prof. Ceni, Andrea
controrelatore Prof. Cossu, Andrea
Keywords
- deep learning
- recurrent neural networks
- reservoir computing
Graduation session start date
11/04/2025
Availability
Full
Abstract (Inglese)
Abstract (Italiano)
Sviluppo del modello deep Reservoir Memory Network (deepRMN), un estensione del modello Reservoir Memory Network (RMN), in ambito di reservoir computing. Nella tesi è stato formalizzato architetturalmente e teoricamente il modello deepRMN, in particolare si sono studiate, dal punto di vista matematico, le condizioni nelle quali le sue dinamiche risultano essere stabili. Sono inoltre stati fatti una serie di tests sperimentali per verificare i punti di forza e di debolezza del modello.
Development of the deep Reservoir Memory Network (deepRMN) model, an extension of the Reservoir Memory Network (RMN) model, in the area of reservoir computing. In the thesis, the deepRMN model was formalized architecturally and theoretically; in particular, the conditions under which its dynamics are found to be stable were studied mathematically. A series of experimental tests were also done to verify the strengths and weaknesses of the model.
Development of the deep Reservoir Memory Network (deepRMN) model, an extension of the Reservoir Memory Network (RMN) model, in the area of reservoir computing. In the thesis, the deepRMN model was formalized architecturally and theoretically; in particular, the conditions under which its dynamics are found to be stable were studied mathematically. A series of experimental tests were also done to verify the strengths and weaknesses of the model.
File
| Nome file | Dimensione |
|---|---|
| Tesi_magistrale.pdf | 2.32 Mb |
Contatta l’autore |
|