Reservoir Computing in MEMS

Dion, Guillaume; Oudrhiri, A. Idrissi-El; Barazani, Bruno; Tessier-Poirier, Albert; Sylvestre, Julien

doi:10.1007/978-981-13-1687-6_9

Cited by 8 publications

(3 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here the goal is to exploit the rich dynamics of complex physical systems as information-processing devices. Physical substrates used for reservoirs are quite diverse: from analog circuits [131][132][133][134], field programmable gate arrays [135][136][137][138], photonic opto-electronic devices [139][140][141][142][143][144], spintronics [145][146][147], quantum dynamics [148,149], nanomaterials [150][151][152][153][154][155][156], biological materials and organoids [157][158][159][160][161][162], mechanics and robotics [163][164][165], up to liquids or fluids [166,167], and most recently, origami structures [168]. As physical reservoir computing becomes more popular, we envision the use of conn2res as a workbench to explore the effect of network interactions on the computational properties of physical reservoirs.…”

Section: Discussionmentioning

confidence: 99%

conn2res: A toolbox for connectome-based reservoir computing

Mišić

Suárez

Mihalik³

et al. 2023

Preprint

View full text Add to dashboard Cite

The connection patterns of neural circuits form a complex network. How signaling in these circuits manifests as complex cognition and adaptive behaviour remains the central question in neuroscience. Concomitant advances in connectomics and artificial intelligence open fundamentally new opportunities to understand how connection patterns shape computational capacity in biological brain networks. Reservoir computing is a versatile paradigm that uses nonlinear dynamics of high-dimensional dynamical systems to perform computations and approximate cognitive functions. Here we present conn2res: an open-source Python toolbox for implementing biological neural networks as artificial neural networks. conn2res is modular, allowing arbitrary architectures and arbitrary dynamics to be imposed. The toolbox allows researchers to input connectomes reconstructed using multiple techniques, from tract tracing to noninvasive diffusion imaging, and to impose multiple dynamical systems, from simple spiking neurons to memristive dynamics. The versatility of the conn2res toolbox allows us to ask new questions at the confluence of neuroscience and artificial intelligence. By reconceptualizing function as computation, conn2res sets the stage for a more mechanistic understanding of structure-function relationships in brain networks.

show abstract

Section: Discussionmentioning

confidence: 99%

conn2res: A toolbox for connectome-based reservoir computing

Mišić

Suárez

Mihalik³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…This has led to a wide variety of different implementations of in materia RC using e.g. photonic [6], mechanical [7] and memristive [8,9] systems.…”

Section: Introductionmentioning

confidence: 99%

Exploring physical and digital architectures in magnetic nanoring array reservoir computers

Venkat,

Vidamour,

Swindells

et al. 2024

Neuromorph. Comput. Eng.

View full text Add to dashboard Cite

Physical reservoir computing (RC) is a machine learning technique that is ideal for processing of time dependent data series. It is also uniquely well-aligned to in materio computing realisations that allow the inherent memory and non-linear responses of functional materials to be directly exploited for computation. We have previously shown that square arrays of interconnected magnetic nanoringsare attractive candidates for in materio reservoir computing, and experimentally demonstrated their strong performance in a range of benchmark tasks. Here, we extend these studies to other lattice arrangements of rings, including trigonal and Kagome grids, to explore how these affect both the magnetic behaviours of the arrays, and their computational properties. We show that while lattice geometry substantially affects the microstate behaviour of the arrays, these differences manifest less profoundly when averaging magnetic behaviour across the arrays. Consequently the computational properties (as measured using task agnostic metrics) of devices with a single electrical readout are found to be only subtly different, with the approach used to time-multiplex data into and out of the arrays having a stronger effect on properties than the lattice geometry. However, we also find that hybrid reservoirs that combine the outputs from arrays with different lattice geometries show enhanced computational properties compared to any single array.

show abstract

“…As the response of the RNN is mathematically analogous to that of a dynamic system, it can be substituted with a real-world dynamic system with appropriate properties, namely nonlinearity between input and output, and a dependence on previous state that asymptotically diminishes over time, termed a 'fading memory'. This has led to a plethora of proposed implementations, with platforms including optoelectronic [5][6][7] , molecular 8 , mechanical [9][10][11] , biological 12,13 , memristive [14][15][16] and magnetic [17][18][19][20][21][22][23] systems.…”

mentioning

confidence: 99%

Reconfigurable reservoir computing in a magnetic metamaterial

Vidamour,

Swindells,

Venkat

et al. 2023

Commun Phys

View full text Add to dashboard Cite

In-materia reservoir computing (RC) leverages the intrinsic physical responses of functional materials to perform complex computational tasks. Magnetic metamaterials are exciting candidates for RC due to their huge state space, nonlinear emergent dynamics, and non-volatile memory. However, to be suitable for a broad range of tasks, the material system is required to exhibit a broad range of properties, and isolating these behaviours experimentally can often prove difficult. By using an electrically accessible device consisting of an array of interconnected magnetic nanorings- a system shown to exhibit complex emergent dynamics- here we show how reconfiguring the reservoir architecture allows exploitation of different aspects the system’s dynamical behaviours. This is evidenced through state-of-the-art performance in diverse benchmark tasks with very different computational requirements, highlighting the additional computational configurability that can be obtained by altering the input/output architecture around the material system.

show abstract

Reservoir Computing in MEMS

Cited by 8 publications

References 56 publications

conn2res: A toolbox for connectome-based reservoir computing

conn2res: A toolbox for connectome-based reservoir computing

Exploring physical and digital architectures in magnetic nanoring array reservoir computers

Reconfigurable reservoir computing in a magnetic metamaterial

Contact Info

Product

Resources

About