Training a multilayer dynamical spintronic network with standard machine-learning tools to perform time-series classification

被引：0

作者：

Plouet, Erwan ^{[1
]}

Sanz-Hernández, Dédalo ^{[1
]}

Vecchiola, Aymeric ^{[1
]}

Grollier, Julie ^{[1
]}

Mizrahi, Frank ^{[1
]}

机构：

[1] Laboratoire Albert Fert, CNRS, Thales, Universite Paris-Saclay, Palaiseau

来源：

Physical Review Applied | 2025年 / 23卷 / 03期

基金：

欧盟地平线“2020”;

关键词：

Multilayer neural networks - Oscillators (electronic) - Oscillators (mechanical) - Spintronics;

D O I：

10.1103/PhysRevApplied.23.034051

中图分类号：

学科分类号：

摘要：

The ability to process time series at low energy cost is critical for many applications. Recurrent neural networks, which can perform such tasks, are computationally expensive when implemented in software on conventional computers. Here we propose to implement a recurrent neural network in hardware using spintronic oscillators as dynamical neurons. Using numerical simulations, we build a multilayer network and demonstrate that we can use back-propagation through time and standard machine-learning tools to train this network. Leveraging the transient dynamics of the spintronic oscillators, we solve the sequential digits classification task with 89.83%±2.91% accuracy, as good as the equivalent software network. We devise guidelines on how to choose the time constant of the oscillators as well as hyperparameters of the network to adapt to different input time scales. © 2025 American Physical Society.

引用

共 44 条

[1] Pineda F., Neural Information Processing Systems, (1987)
[2] Lecun Y., Artificial Neural Networks: Concepts and Theory, (1992)
[3] Hermans M., Schrauwen B., Advances in Neural Information Processing Systems, 26, (2013)
[4] Hasani R., Lechner M., Amini A., Rus D., Grosu R., Proceedings of the AAAI Conference on Artificial Intelligence, 35, pp. 7657-7666, (2021)
[5] Eshraghian J. K., Ward M., Neftci E. O., Wang X., Lenz G., Dwivedi G., Bennamoun M., Jeong D. S., Lu W. D., Training spiking neural networks using lessons from deep learning, Proc. IEEE, 111, (2023)
[6] Taherkhani A., Belatreche A., Li Y., Cosma G., Maguire L. P., McGinnity T. M., A review of learning in biologically plausible spiking neural networks, Neural Netw, 122, (2020)
[7] Tavanaei A., Ghodrati M., Kheradpisheh S. R., Masquelier T., Maida A., Deep learning in spiking neural networks, Neural Netw, 111, (2019)
[8] Yin B., Corradi F., Bohte S. M., Accurate and efficient time-domain classification with adaptive spiking recurrent neural networks, Nat. Mach. Intell, 3, (2021)
[9] Chen R. T., Rubanova Y., Bettencourt J., Duvenaud D. K., Advances in Neural Information Processing Systems, 31, (2018)
[10] Pehle C., Billaudelle S., Cramer B., Kaiser J., Schreiber K., Stradmann Y., Weis J., Leibfried A., Muller E., Schemmel J., The BrainScales-2 accelerated neuromorphic system with hybrid plasticity, Front. Neurosci, 16, (2022)

← 1 2 3 4 5 →