Physical reservoir computing is a brain-inspired computational framework that allows information processing by exploiting the complex dynamics of high-dimensional physical systems. In this context, self-organizing memristive systems composed of interacting nano-objects have been proposed as multipurpose platforms for the hardware implementation of reservoir computing (RC). Here, we report on Mackey-Glass time series prediction with memristive nanowire (NW) networks. Besides showing that emergent memristive dynamics of these networks modeled through a graph theoretical approach can be exploited for time series prediction, it is shown that the accuracy of the system can be tailored through appropriate configuration of the multiterminal NW network. Results show that NW networks can be exploited for in materia implementation of reservoir computing paradigm towards the realization of brain-inspired neuromorphic systems based on low-cost self-organizing nanomaterials.
Mackey-Glass Time Series Forecasting by Nanowire Networks / Milano, G.; Chakrabarty, T.; Ricciardi, C.. - (2023), pp. 989-994. (Intervento presentato al convegno 2nd Edition IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2023 tenutosi a Milan (ITA) nel 25-27 October 2023) [10.1109/MetroXRAINE58569.2023.10405786].
Mackey-Glass Time Series Forecasting by Nanowire Networks
Milano G.;Ricciardi C.
2023
Abstract
Physical reservoir computing is a brain-inspired computational framework that allows information processing by exploiting the complex dynamics of high-dimensional physical systems. In this context, self-organizing memristive systems composed of interacting nano-objects have been proposed as multipurpose platforms for the hardware implementation of reservoir computing (RC). Here, we report on Mackey-Glass time series prediction with memristive nanowire (NW) networks. Besides showing that emergent memristive dynamics of these networks modeled through a graph theoretical approach can be exploited for time series prediction, it is shown that the accuracy of the system can be tailored through appropriate configuration of the multiterminal NW network. Results show that NW networks can be exploited for in materia implementation of reservoir computing paradigm towards the realization of brain-inspired neuromorphic systems based on low-cost self-organizing nanomaterials.File | Dimensione | Formato | |
---|---|---|---|
2023_IEEE_Mackey-Glass.pdf
accesso riservato
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Non Pubblico - Accesso privato/ristretto
Dimensione
1.16 MB
Formato
Adobe PDF
|
1.16 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2995532