Reservoir computing is a paradigm in machine learning whose processing capabilities rely on the dynamical behavior of recurrent neural networks. We present a mixed analog and digital implementation of this concept with a nonlinear analog electronic circuit as a main computational unit. In our approach, the reservoir network can be replaced by a single nonlinear element with delay via time-multiplexing. We analyze the influence of noise on the performance of the system for two benchmark tasks: 1) a classification problem and 2) a chaotic time-series prediction task. Special attention is given to the role of quantization noise, which is studied by varying the resolution in the conversion interface between the analog and digital worlds.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TNNLS.2014.2311855 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
IMPACT: In-Memory ComPuting Architecture based on Y-FlAsh Technology for Coalesced Tsetlin machine inference.
Philos Trans A Math Phys Eng Sci
January 2025
Microsystems Group, School of Engineering, Newcastle University, Newcastle upon Tyne NE1 7RU, UK.
The increasing demand for processing large volumes of data for machine learning (ML) models has pushed data bandwidth requirements beyond the capability of traditional von Neumann architecture. In-memory computing (IMC) has recently emerged as a promising solution to address this gap by enabling distributed data storage and processing at the micro-architectural level, significantly reducing both latency and energy. In this article, we present In-Memory comPuting architecture based on Y-FlAsh technology for Coalesced Tsetlin machine inference (IMPACT), underpinned on a cutting-edge memory device, Y-Flash, fabricated on a 180 nm complementary metal oxide semiconductor (CMOS) process.
View Article and Find Full Text PDFDual-Modal Memory Enabled by a Single Vertical N-Type Organic Artificial Synapse for Neuromorphic Computing.
ACS Appl Mater Interfaces
January 2025
School of Materials and Energy, Lanzhou University (LZU), Lanzhou 730000, China.
Complementary neural network circuits combining multifunctional high-performance p-type with n-type organic artificial synapses satisfy sophisticated applications such as image cognition and prosthesis control. However, implementing the dual-modal memory features that are both volatile and nonvolatile in a synaptic transistor is challenging. Herein, for the first time, we propose a single vertical n-type organic synaptic transistor (VNOST) with a novel polymeric organic mixed ionic-electronic conductor as the core channel material to achieve dual-modal synaptic learning/memory behaviors at different operating current densities via the formation of an electric double layer and the reversible ion doping.
View Article and Find Full Text PDFBond strength to eroded dentin as per chlorhexidine use for controlling erosive wear or interface aging: an 18-month assay.
Braz Oral Res
January 2025
Universidade de São Paulo - USP, School of Dentistry, Department of Operative Dentistry, São Paulo, SP, Brazil.
The aim of this study was to assess the effect of a chlorhexidine digluconate solution (CHX) applied as an antiproteolytic agent for controlling erosive tooth wear or as part of the adhesive treatment on long-term bond strength to eroded dentin. Dentin specimens were abraded with a 600-grit silicon carbide (SiC) paper for 1 min (sound dentin - S), subsequently treated with 2% CHX for 1 min (with excess removed, followed by a 6-hour rest), and eroded by exposure to Coca-Cola for 5 min, three times a day, for 5 days (CHX-treated and eroded dentin - CHXE), or only eroded (eroded dentin - E). The specimens were acid-etched (15 s), rinsed (30 s), dried (15 s), and rehydrated with 1.
View Article and Find Full Text PDFsp. nov., a bacterium isolated from the roots of the aquatic plant .
Int J Syst Evol Microbiol
January 2025
Department of Biology, Slippery Rock University, Slippery Rock, Pennsylvania 16057, USA.
A polyphasic taxonomic study was carried out on strain T9W2-O, isolated from the roots of the aquatic plant . This isolate is rod-shaped, forms yellow/orange pigmented colonies and produces the pigment flexirubin. Nearly complete 16S rRNA gene sequence homology related the strain to , with 98.
View Article and Find Full Text PDFEfficacy of virtual reality balance training on rehabilitation outcomes following anterior cruciate ligament reconstruction: A systematic review and meta-analysis.
PLoS One
January 2025
Orthopedics Department, First Teaching Hospital of Tianjin University of Traditional Chinese Medicine, Tianjin, China.
Objective: The objective of this systematic review and meta-analysis is to clarify the rehabilitation efficacy of virtual reality (VR) balance training after anterior cruciate ligament reconstruction (ACLR).
Methods: This meta-analysis was registered in PROSPERO with the registration number CRD42024520383. The electronic databases PubMed, Web of Science, Cochrane Library, MEDLINE, Embase, China National Knowledge Infrastructure, Chinese Biomedical Literature, China Science and Technology Journal Database, and Wanfang Digital Periodical database were systematically searched to identify eligible studies from their inception up to January 2024.
Want 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!