Dynamic memristor-based reservoir computing for high-efficiency temporal signal processing.

Reservoir computing is a highly efficient network for processing temporal signals due to its low training cost compared to standard recurrent neural networks, and generating rich reservoir states is critical in the hardware implementation. In this work, we report a parallel dynamic memristor-based r... View More