\n A pruning reservoir computing technique can mitigate noise and reconstruct nonlinear dynamics for potential engineering and neuroscience applications.\n <\/p>\n
Many systems in the world are nonlinear: Changes in their input are not proportional to changes in their output. But scientists and engineers want to predict the behavior of these complex nonlinear systems \u2014 such as neuron firing \u2014 to better understand and manage them. One challenge is reconstructing unobserved states from limited, noisy measurements of nonlinear systems. While filtering techniques are often used to mitigate noise, they require the inclusion of possibly unknown equations that govern the underlying physics processes.<\/p>\n
A promising alternative is reservoir computing, an artificial intelligence and machine learning technique that efficiently simulates systems using a neural network with random connectivity, known as the reservoir. Sedehi et al. present a truncated reservoir computing approach that can distinguish noise from signal and reconstruct nonlinear dynamics from noisy, incomplete data without using any physics-based models.<\/p>\n
The approach enhances the efficiency of the reservoir by pruning redundant nodes and edges \u2014 a form of data compression \u2014 and optimizing hyperparameters with a novel machine learning protocol. Applying the approach to two nonlinear systems, the Lorenz attractor and a simplified model of biological neuron firing, the authors found its ability to distinguish noise from signal competes with conventional filtering techniques at low signal-to-noise ratios and in high-frequency ranges.<\/p>\n
\u201cThis research paves the way for model-free denoising and state reconstruction in complex systems with reservoir computing,\u201d said author Omid Sedehi. \u201cThe proposed framework is adaptable for hardware implementations, making it highly relevant for new applications in engineering, neuroscience, bioacoustics, and physical computation.\u201d<\/p>\n
Future work will focus on scaling the pruning technique to deal with high-dimensional reservoirs and provide hardware-ready implementations for practical deployment.<\/p>\n
Source: \u201cDenoising and reconstruction of nonlinear dynamics using truncated reservoir computing,\u201d by Omid Sedehi, Manish Yadav, Merten Stender, and Sebastian Oberst, Chaos (2025). The article can be accessed at https:\/\/doi.org\/10.1063\/5.0273505 <\/a>.<\/p>\n