Linear Simple Cycle Reservoirs at the edge of stability perform Fourier decomposition of the input driving signals
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866913596324184064 |
|---|---|
| author | Fong, Robert Simon Li, Boyu Tino, Peter |
| author_facet | Fong, Robert Simon Li, Boyu Tino, Peter |
| contents | This paper explores the representational structure of linear Simple Cycle Reservoirs (SCR) operating at the edge of stability. We view SCR as providing in their state space feature representations of the input-driving time series. By endowing the state space with the canonical dot-product, we ``reverse engineer" the corresponding kernel (inner product) operating in the original time series space. The action of this time-series kernel is fully characterized by the eigenspace of the corresponding metric tensor. We demonstrate that when linear SCRs are constructed at the edge of stability, the eigenvectors of the time-series kernel align with the Fourier basis. This theoretical insight is supported by numerical experiments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_00295 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Linear Simple Cycle Reservoirs at the edge of stability perform Fourier decomposition of the input driving signals Fong, Robert Simon Li, Boyu Tino, Peter Neural and Evolutionary Computing Dynamical Systems This paper explores the representational structure of linear Simple Cycle Reservoirs (SCR) operating at the edge of stability. We view SCR as providing in their state space feature representations of the input-driving time series. By endowing the state space with the canonical dot-product, we ``reverse engineer" the corresponding kernel (inner product) operating in the original time series space. The action of this time-series kernel is fully characterized by the eigenspace of the corresponding metric tensor. We demonstrate that when linear SCRs are constructed at the edge of stability, the eigenvectors of the time-series kernel align with the Fourier basis. This theoretical insight is supported by numerical experiments. |
| title | Linear Simple Cycle Reservoirs at the edge of stability perform Fourier decomposition of the input driving signals |
| topic | Neural and Evolutionary Computing Dynamical Systems |
| url | https://arxiv.org/abs/2412.00295 |