Tag Archives: reservoir

NEW PAPER: Spoken Digit Classification by In-Materio Reservoir Computing With Neuromorphic Atomic Switch Networks

Lilak S, Woods W, Scharnhorst K, Dunham C, Teuscher C, Stieg AZ and Gimzewski JK (2021) Spoken Digit Classification by In-Materio Reservoir Computing With Neuromorphic Atomic Switch Networks. Frontiers in Nanotechnology, 3:675792. doi: 10.3389/fnano.2021.675792

Abstract: Atomic Switch Networks comprising silver iodide (AgI) junctions, a material previously unexplored as functional memristive elements within highly interconnected nanowire networks, were employed as a neuromorphic substrate for physical Reservoir Computing. This new class of ASN-based devices has been physically characterized and utilized to classify spoken digit audio data, demonstrating the utility of substrate-based device architectures where intrinsic material properties can be exploited to perform computation in-materio. This work demonstrates high accuracy in the classification of temporally analyzed Free-Spoken Digit Data These results expand upon the class of viable memristive materials available for the production of functional nanowire networks and bolster the utility of ASN-based devices as unique hardware platforms for neuromorphic computing applications involving memory, adaptation and learning.

New paper: Deep reservoir computing with memcapacitors

S. J. Dat Tran and C. Teuscher, “Deep Memcapacitive Network,” 2020 IEEE 15th International Conference on Nano/Micro Engineered and Molecular System (NEMS), San Diego, CA, USA, 2020, pp. 200-205, doi: https://doi.org/10.1109/NEMS50311.2020.9265561

 

Alife 2020 paper

Our Alife 2020 paper was accepted for publication.

H. Nguyen, P. Banda, D. Stefanovic, and C. Teuscher, Reservoir Computing with Random Chemical Systems, Proceedings of The 2020 Conference on Artificial Life, MIT Press, pp. 491-499, 2020.  https://doi.org/10.1162/isal_a_00324

Abstract: Top-down engineering of biomolecular circuits to perform specific computational tasks is notoriously hard and time-consuming. Current circuits have limited complexity and are brittle and application-specific. Here we propose an alternative: we design and test a bottom-up constructed Reservoir Computer (RC) that uses random chemical circuits inspired by DNA strand displacement reactions. This RC has the potential to be implemented easily and trained for various tasks. We describe and simulate it by means of a Chemical Reaction Network (CRN) and evaluate its performance on three computational tasks: the Hamming distance and a short- as well as a long-term memory. Compared with the deoxyribozyme oscillator RC model simulated by Yahiro et al., our random chemical RC performs 75.5% better for the short-term and 67.2% better for the long-term memory task. Our model requires an 88.5% larger variety of chemical species, but it relies on random chemical circuits, which can be more easily realized and scaled up. Thus, our novel random chemical RC has the potential to simplify the way we build adaptive biomolecular circuits.

Two New Reservoir Computing Papers Published

N. Babson and C. Teuscher, Reservoir Computing with Complex Cellular Automata, Complex Systems, 28(4), 2019 pp. 433–455.
https://doi.org/10.25088/ComplexSystems.28.4.433

S. J. D. Tran and C. Teuscher, “Hierarchical Memcapacitive Reservoir Computing Architecture,” 2019 IEEE International Conference on Rebooting Computing (ICRC), San Mateo, CA, USA, 2019, pp. 1-6. https://doi.org/10.1109/ICRC.2019.8914716