In linear reservoir computing, dominant training modes approximate the Koopman eigenfunctions of the data-generating dynamical system, linking RC directly to DMD.
On explaining the surprising success of reservoir computing forecaster of chaos? the universal machine learning dynamical system with contrast to var and dmd
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
On Dominant Manifolds in Reservoir Computing Networks
In linear reservoir computing, dominant training modes approximate the Koopman eigenfunctions of the data-generating dynamical system, linking RC directly to DMD.