Reservoir Computing in Materio: An Evaluation of Configuration through Evolution

Matthew Nicholas Dale; Susan Stepney; Julian Francis Miller; Martin Albrecht Trefzer

doi:10.1109/SSCI.2016.7850170

Reservoir Computing in Materio: An Evaluation of Configuration through Evolution

Matthew Nicholas Dale, Susan Stepney, Julian Francis Miller, Martin Albrecht Trefzer

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

Abstract

Recent work has shown that computational substrates made from carbon nanotube/polymer mixtures can form trainable Reservoir Computers. This new reservoir computing platform uses computer based evolutionary algorithms to optimise a set of electrical control signals to induce reservoir properties within the substrate. In the training process, evolution decides the value of analogue control signals (voltages) and the location of inputs and outputs on the substrate that improve the performance of the subsequently trained reservoir readout.
Here, we evaluate the performance of evolutionary search compared to randomly assigned electrical configurations. The substrate is trained and evaluated on time-series prediction using the Santa Fe Laser generated competition data (dataset A). In addition to the main investigation, we introduce two new features closely linked to the traditional reservoir computing architecture, adding an evolvable input-weighting mechanism and a reservoir time-scaling parameter.
The experimental results show evolved configurations across all four test substrates consistently produce reservoirs with greater performance than randomly configured reservoirs. The results also show that applying both input-weighting and timescaling simultaneously can provide additional tuning to the task, improving performance. For one material, the evolved reservoir is shown to outperform – for this task – all other hardwarebased reservoir computers found in the literature. The same material also outperforms a simple evolved simulated Echo State Network of the same size. The performance of this material is reported to be both consistent after long time-periods and after reconfiguration to other tasks.

Original language	English
Title of host publication	IEEE Symposium Series on Computational Intelligence (SSCI), 2016
Publisher	IEEE
DOIs	https://doi.org/10.1109/SSCI.2016.7850170
Publication status	Published - 2016

Bibliographical note

This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

Access to Document

10.1109/SSCI.2016.7850170

ices-paper-2016
This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details
Accepted author manuscript, 960 KB

Cite this

@inbook{a70cb8ef7b854da68bb84453f2cee846,

title = "Reservoir Computing in Materio: An Evaluation of Configuration through Evolution",

abstract = "Recent work has shown that computational substrates made from carbon nanotube/polymer mixtures can form trainable Reservoir Computers. This new reservoir computing platform uses computer based evolutionary algorithms to optimise a set of electrical control signals to induce reservoir properties within the substrate. In the training process, evolution decides the value of analogue control signals (voltages) and the location of inputs and outputs on the substrate that improve the performance of the subsequently trained reservoir readout. Here, we evaluate the performance of evolutionary search compared to randomly assigned electrical configurations. The substrate is trained and evaluated on time-series prediction using the Santa Fe Laser generated competition data (dataset A). In addition to the main investigation, we introduce two new features closely linked to the traditional reservoir computing architecture, adding an evolvable input-weighting mechanism and a reservoir time-scaling parameter. The experimental results show evolved configurations across all four test substrates consistently produce reservoirs with greater performance than randomly configured reservoirs. The results also show that applying both input-weighting and timescaling simultaneously can provide additional tuning to the task, improving performance. For one material, the evolved reservoir is shown to outperform – for this task – all other hardwarebased reservoir computers found in the literature. The same material also outperforms a simple evolved simulated Echo State Network of the same size. The performance of this material is reported to be both consistent after long time-periods and after reconfiguration to other tasks.",

author = "Dale, {Matthew Nicholas} and Susan Stepney and Miller, {Julian Francis} and Trefzer, {Martin Albrecht}",

note = "This is an author-produced version of the published paper. Uploaded in accordance with the publisher{\textquoteright}s self-archiving policy. Further copying may not be permitted; contact the publisher for details",

year = "2016",

doi = "10.1109/SSCI.2016.7850170",

language = "English",

booktitle = "IEEE Symposium Series on Computational Intelligence (SSCI), 2016",

publisher = "IEEE",

address = "United States",

}

TY - CHAP

T1 - Reservoir Computing in Materio

T2 - An Evaluation of Configuration through Evolution

AU - Dale, Matthew Nicholas

AU - Stepney, Susan

AU - Miller, Julian Francis

AU - Trefzer, Martin Albrecht

N1 - This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

PY - 2016

Y1 - 2016

N2 - Recent work has shown that computational substrates made from carbon nanotube/polymer mixtures can form trainable Reservoir Computers. This new reservoir computing platform uses computer based evolutionary algorithms to optimise a set of electrical control signals to induce reservoir properties within the substrate. In the training process, evolution decides the value of analogue control signals (voltages) and the location of inputs and outputs on the substrate that improve the performance of the subsequently trained reservoir readout. Here, we evaluate the performance of evolutionary search compared to randomly assigned electrical configurations. The substrate is trained and evaluated on time-series prediction using the Santa Fe Laser generated competition data (dataset A). In addition to the main investigation, we introduce two new features closely linked to the traditional reservoir computing architecture, adding an evolvable input-weighting mechanism and a reservoir time-scaling parameter. The experimental results show evolved configurations across all four test substrates consistently produce reservoirs with greater performance than randomly configured reservoirs. The results also show that applying both input-weighting and timescaling simultaneously can provide additional tuning to the task, improving performance. For one material, the evolved reservoir is shown to outperform – for this task – all other hardwarebased reservoir computers found in the literature. The same material also outperforms a simple evolved simulated Echo State Network of the same size. The performance of this material is reported to be both consistent after long time-periods and after reconfiguration to other tasks.

AB - Recent work has shown that computational substrates made from carbon nanotube/polymer mixtures can form trainable Reservoir Computers. This new reservoir computing platform uses computer based evolutionary algorithms to optimise a set of electrical control signals to induce reservoir properties within the substrate. In the training process, evolution decides the value of analogue control signals (voltages) and the location of inputs and outputs on the substrate that improve the performance of the subsequently trained reservoir readout. Here, we evaluate the performance of evolutionary search compared to randomly assigned electrical configurations. The substrate is trained and evaluated on time-series prediction using the Santa Fe Laser generated competition data (dataset A). In addition to the main investigation, we introduce two new features closely linked to the traditional reservoir computing architecture, adding an evolvable input-weighting mechanism and a reservoir time-scaling parameter. The experimental results show evolved configurations across all four test substrates consistently produce reservoirs with greater performance than randomly configured reservoirs. The results also show that applying both input-weighting and timescaling simultaneously can provide additional tuning to the task, improving performance. For one material, the evolved reservoir is shown to outperform – for this task – all other hardwarebased reservoir computers found in the literature. The same material also outperforms a simple evolved simulated Echo State Network of the same size. The performance of this material is reported to be both consistent after long time-periods and after reconfiguration to other tasks.

U2 - 10.1109/SSCI.2016.7850170

DO - 10.1109/SSCI.2016.7850170

M3 - Chapter (peer-reviewed)

BT - IEEE Symposium Series on Computational Intelligence (SSCI), 2016

PB - IEEE

ER -