Projects per year
Abstract
Reservoir Computing (RC) models, a subclass of recurrent neural networks, are distinguished by their fixed, non-trainable input layer and dynamically coupled reservoir, with only the static readout layer being trained. This design circumvents the issues associated with backpropagating error signals through time, thereby enhancing both stability and training efficiency. RC models have been successfully applied across a broad range of application domains. Crucially, they have been demonstrated to be universal approximators of time-invariant dynamic filters with fading memory, under various settings of approximation norms and input driving sources. Simple Cycle Reservoirs (SCR) represent a specialized class of RC models with a highly constrained reservoir architecture, characterized by uniform ring connectivity and binary input-to-reservoir weights with an aperiodic sign pattern. For linear reservoirs, given the reservoir size, the reservoir construction has only one degree of freedom -- the reservoir cycle weight. Such architectures are particularly amenable to hardware implementations without significant performance degradation in many practical tasks. In this study we endow these observations with solid theoretical foundations by proving that SCRs operating in real domain are universal approximators of time-invariant dynamic filters with fading memory. Our results supplement recent research showing that SCRs in the complex domain can approximate, to arbitrary precision, any unrestricted linear reservoir with a non-linear readout. We furthermore introduce a novel method to drastically reduce the number of SCR units, making such highly constrained architectures natural candidates for low-complexity hardware implementations. Our findings are supported by empirical studies on real-world time series datasets.
| Original language | English |
|---|---|
| Title of host publication | AAAI-25 Technical Tracks 16 |
| Subtitle of host publication | Thirty-Ninth AAAI Conference on Artificial Intelligence, Thirty-Seventh Conference on Innovative Applications of Artificial Intelligence, Fifteenth Symposium on Educational Advances in Artificial Intelligence |
| Editors | Toby Walsh, Julie Shah, Zico Kolter |
| Place of Publication | Washington, DC |
| Publisher | AAAI Press |
| Pages | 16622-16629 |
| Number of pages | 8 |
| Edition | 1 |
| ISBN (Print) | 9781577358978 |
| DOIs | |
| Publication status | Published - 11 Apr 2025 |
| Event | 39th AAAI Conference on Artificial Intelligence - Philadelphia Marriott Downtown, Philadelphia, United States Duration: 25 Feb 2025 → 4 Mar 2025 https://aaai.org/conference/aaai/aaai-25/ |
Publication series
| Name | Proceedings of the AAAI Conference on Artificial Intelligence |
|---|---|
| Publisher | AAAI |
| Number | 16 |
| Volume | 39 |
| ISSN (Print) | 2374-3468 |
| ISSN (Electronic) | 2374-3468 |
Conference
| Conference | 39th AAAI Conference on Artificial Intelligence |
|---|---|
| Abbreviated title | AAAI-2025 |
| Country/Territory | United States |
| City | Philadelphia |
| Period | 25/02/25 → 4/03/25 |
| Internet address |
Fingerprint
Dive into the research topics of 'Universality of Real Minimal Complexity Reservoir'. Together they form a unique fingerprint.Projects
- 1 Active
-
Advanced Research into Crystallographic Anisotropy & Nucleation Effects in single crystals (ARCANE)
Withey, P. (Co-Investigator), Bowen, P. (Co-Investigator), Styles, I. (Co-Investigator), Green, N. (Principal Investigator), Tino, P. (Co-Investigator) & Warnken, N. (Co-Investigator)
Rolls Royce, Engineering & Physical Science Research Council
1/06/23 → 31/05/28
Project: Research Councils
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver