Reservoir Computing Techniques Using Tensor Networks

Shinji Sato, Daiki Sasaki, Chih-Chieh Chen, Kodai Shiba, Tomah Sogabe

Abstract


Reservoir computing (RC), traditionally based on echo state networks and liquid state machines, shows great potential in modeling dynamic time-series data like weather and astronomical predictions. However, these frameworks are not suitable for quantum dynamics-based RC. Tensor networks (TNs) are well suited for modeling quantum dynamics because they can efficiently model quantum information and entanglement. In this work, we propose a novel randomized TN-based RC scheme, demonstrating its validity through various case studies. Our results show superior performance compared to traditional ESN models, laying the groundwork for further exploration of quantum reservoir computing.

Keywords


Reservoir Computing (RC); Tensor Networks (TN); Quantum Dynamics

Full Text:

PDF

Refbacks

  • There are currently no refbacks.