Learning strange attractors with reservoir systems
This work offers foundational insights for representing and analyzing chaotic systems, with implications for machine learning in dynamical systems.
The paper generalizes Takens' Embedding Theorem, showing that random linear state-space representations of observations from invertible dynamical systems embed phase space dynamics, providing tools for learning chaotic attractors and explaining reservoir computing in recurrent neural networks.
This paper shows that the celebrated Embedding Theorem of Takens is a particular case of a much more general statement according to which, randomly generated linear state-space representations of generic observations of an invertible dynamical system carry in their wake an embedding of the phase space dynamics into the chosen Euclidean state space. This embedding coincides with a natural generalized synchronization that arises in this setup and that yields a topological conjugacy between the state-space dynamics driven by the generic observations of the dynamical system and the dynamical system itself. This result provides additional tools for the representation, learning, and analysis of chaotic attractors and sheds additional light on the reservoir computing phenomenon that appears in the context of recurrent neural networks.