We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.ET

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Emerging Technologies

Title: Reservoir computing for sensing: an experimental approach

Abstract: The increasing popularity of machine learning solutions puts increasing restrictions on this field if it is to penetrate more aspects of life. In particular, energy efficiency and speed of operation is crucial, inter alia in portable medical devices. The Reservoir Computing (RC) paradigm poses as a solution to these issues through foundation of its operation: the reservoir of states. Adequate separation of input information translated into the internal state of the reservoir, whose connections do not need to be trained, allow to simplify the readout layer thus significantly accelerating the operation of the system. In this brief review article, the theoretical basis of RC was first described, followed by a description of its individual variants, their development and state-of-the-art applications in chemical sensing and metrology: detection of impedance changes and ion sensing. Presented results indicate applicability of reservoir computing for sensing and validating the SWEET algorithm experimentally.
Subjects: Emerging Technologies (cs.ET)
Journal reference: International Journal of Unconventional Computing, 2019, 14, 267
Cite as: arXiv:2001.04342 [cs.ET]
  (or arXiv:2001.04342v1 [cs.ET] for this version)

Submission history

From: Konrad Szaciłowski [view email]
[v1] Fri, 10 Jan 2020 15:34:52 GMT (897kb)

Link back to: arXiv, form interface, contact.