Connecting reservoir computing with statistical forecasting and deep neural networks

Among the existing machine learning frameworks, reservoir computing demonstrates fast and low-cost training, and its suitability for implementation in various physical systems. This Comment reports on how aspects of reservoir computing can be applied to classical forecasting methods to accelerate the learning process, and highlights a new approach that makes the hardware implementation of traditional machine learning algorithms practicable in electronic and photonic systems.

Automated summary

Reservoir computing is a fast and low-cost training method suitable for various physical systems, which can accelerate the learning process. A new approach enables the hardware implementation of traditional machine learning algorithms in electronic and photonic systems.