Skip to content

Scalability of multifunctional reservoir computing

Home / Open Questions / Scalability of multifunctional reservoir computing

Background: Reservoir computing machines can be designed to infer multiple distinct chaotic systems by incorporating a system-label channel, though performance often suffers due to retrieval unreliability and distorted dynamics as the number of stored systems increases.

Question / Future Work: The scalability and performance impact of increasing the number of target chaotic systems beyond two remains an open, unquantified issue in multifunctional reservoir computing (MFRC) schemes. A detailed analysis is required to understand how scaling affects the accuracy of attractor reproduction, the reliability of state retrieval, and the quality of reconstructed bifurcation diagrams as the number of stored dynamics grows.

Why It Matters: Understanding scalability limits is critical for transitioning MFRC from laboratory demonstrations to practical, real-world applications involving complex systems with many potential dynamical states.

Evidence: Although it is generally expected that the machine performance will degrade as the number of target systems increases, a detailed analysis is still needed to quantify the impact of this increase on the accuracy and reliability of the replicated attractors, as well as on the quality of the reconstructed bifurcation diagrams.

Metadata & Links