- Poster presentation
- Open Access
Extreme sensitivity of reservoir computing to small network disruptions
- Philippe Vincent-Lamarre1Email author,
- Guillaume Lajoie2, 3 and
- Jean-Philippe Thivierge1
https://doi.org/10.1186/1471-2202-16-S1-P256
© Vincent-Lamarre et al. 2015
- Published: 18 December 2015
Keywords
- Single Neuron
- Sinusoidal Signal
- Neural Code
- Random Performance
- Chaotic Regime
Performance of damaged reservoirs of 1,000 neurons with FORCE and innate learning algorithms. A. Target signal (green, perfectly replicated with the originally trained network) and the trace of the same network after the removal of one neuron in its reservoir. B. Ten trials (red) with different initial conditions of a damaged network (N-2 neurons) that is trained to peak at 1,000 ms (green) using innate learning. C. Average lag between the target and the output timing (100 trials per condition) as a function of the number of removed neurons. D. Mean squared error as a function of the number of removed neurons.
Declarations
Acknowledgements
This research was funded by grants to J.P.T. from NSERC Discovery and CIHR operating funds.
Authors’ Affiliations
References
- Sussillo D, Abbott LF: Generating Coherent Patterns of Activity from Chaotic Neural Networks. Neuron. 2009, 63: 544-557.PubMedPubMed CentralView ArticleGoogle Scholar
- Laje R, Buonomano DV: Robust timing and motor patterns by taming chaos in recurrent neural networks. Nat Neurosci. 2013, 16: 925-933.PubMedView ArticleGoogle Scholar
- Barak O, Sussillo D, Romo R, Tsodyks M, Abbott LF: From fixed points to chaos: Three models of delayed discrimination. Progress in Neurobiology. 2013, 103: 214-222.PubMedPubMed CentralView ArticleGoogle Scholar
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.