Determination of the Edge of Criticality in Echo State Networks Through Fisher Information Maximization
- Submitting institution
-
University of Exeter
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1781
- Type
- D - Journal article
- DOI
-
10.1109/TNNLS.2016.2644268
- Title of journal
- IEEE Transactions on Neural Networks and Learning Systems
- Article number
- -
- First page
- 706
- Volume
- 29
- Issue
- 3
- ISSN
- 2162-237X
- Open access status
- Compliant
- Month of publication
- January
- Year of publication
- 2017
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
2
- Research group(s)
-
-
- Citation count
- 16
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- The idea of transitions between operating regimes, i.e. between an ordered and chaotic regime, has always been associated with autonomous dynamical systems.
In this paper, we have provided the first method that takes into account the fact that recurrent networks are driven dynamical systems.
The ideas described in this paper allowed us to yield light on the complexity behind the changes in behaviour occurring in input-driven recurrent neural networks. This work has been well received by the community (e.g 10.1016/j.pneurobio.2017.07.002 and 10.1007/s12559-017-9450-z) and has led to several more papers with colleagues (e.g. https://doi.org/10.1016/j.physd.2020.132609)
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -