Reservoir computing (RC) by supervised training, a bio-inspired paradigm, is gaining popularity for processing time-dependent data. Compared to conventional recurrent neural networks, RC is facilely implemented by available hardware and overcomes some obstacles in training period, such as slow convergence and local optimum. In this paper, we propose and characterize a novel reservoir computing system based on a semiconductor laser with double optoelectronic feedback loops. This system shows obvious improvement on prediction, speech recognition and nonlinear channel equalization compared to the traditional reservoir computing systems with single feedback loop. Then some influencing factors to optimize the performance of the new RC are numerically studied, and its great potential of addressing more complex and troubling problems in information processing is expected to be exploited.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1364/OE.27.027431 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Post-processing methods for delay embedding and feature scaling of reservoir computers.
Commun Eng
January 2025
Institut für Physik, Technische Universität Ilmenau, Ilmenau, Germany.
Reservoir computing is a machine learning method that is well-suited for complex time series prediction tasks. Both delay embedding and the projection of input data into a higher-dimensional space play important roles in enabling accurate predictions. We establish simple post-processing methods that train on past node states at uniformly or randomly-delayed timeshifts.
View Article and Find Full Text PDFIdentification of gas outburst precursors based on outburst percolation theory.
Sci Rep
January 2025
School of Mechanics and Engineering, Liaoning Technical University, Fuxin, 123000, China.
Uniaxial compression experiments were conducted on coal rock utilizing a computed tomography (CT) scanning system for real-time monitoring to explain the issue of gas volume significantly exceeding reservoir capacity during coal and gas outbursts. A percolation factor a which can make a significant contribution to the research on premonitory information of gas outbursts is introduced to determine whether percolation occurs in coal rock, and supports the outburst percolation theory. It was found that percolation probability and correlation length increase with greater porosity, and that the number of pore clusters decreases as porosity increases.
View Article and Find Full Text PDFCO flooding plays a crucial role in enhancing oil recovery and achieving carbon reduction targets, particularly in unconventional reservoirs with complex pore structures. The phase behavior of CO and hydrocarbons at different scales significantly affects oil recovery efficiency, yet its underlying mechanisms remain insufficiently understood. This study improves existing thermodynamic models by introducing Helmholtz free energy as a convergence criterion and incorporating adsorption effects in micro- and nano-scale pores.
View Article and Find Full Text PDFAn Immunocytochemistry Method to Investigate the Translationally Active HIV Reservoir.
Int J Mol Sci
January 2025
MRL, Merck & Co., Inc., Rahway, NJ 07065, USA.
Despite the success of combination antiretroviral therapy (cART) to suppress HIV replication, HIV persists in a long-lived reservoir that can give rise to rebounding viremia upon cART cessation. The translationally active reservoir consists of HIV-infected cells that continue to produce viral proteins even in the presence of cART. These active reservoir cells are implicated in the resultant viremia upon cART cessation and likely contribute to chronic immune activation in people living with HIV (PLWH) on cART.
View Article and Find Full Text PDFStructure-guided discovery of viral proteins that inhibit host immunity.
Cell
January 2025
Department of Molecular Genetics, Weizmann Institute of Science, Rehovot 7610001, Israel. Electronic address:
Viruses encode proteins that inhibit host defenses, but sifting through the millions of available viral sequences for immune-modulatory proteins has been so far impractical. Here, we develop a process to systematically screen virus-encoded proteins for inhibitors that physically bind host immune proteins. Focusing on Thoeris and CBASS, bacterial defense systems that are the ancestors of eukaryotic Toll/interleukin-1 receptor (TIR) and cyclic GMP-AMP synthase (cGAS) immunity, we discover seven families of Thoeris and CBASS inhibitors, encompassing thousands of genes widespread in phages.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!