Quantum reservoir processor
One of the key challenges in the realization of neural networks is the simultaneous control of the many weight connections between neurons. Even though techniques are known for the training of such weights in principle, the control of the weights can be complicated to realize with hardware, especially when considering quantum systems. In the classical domain, reservoir computing offers a particularly simple network architecture, where the majority of the network is based on fixed connections with random strength. Training is performed only on a single readout layer, which makes for particularly straightforward training and a geometry potentially more accessible in hardware. Quantum reservoir processing was recently introduced as a theoretical concept [1], where surprisingly small networks of coupled quantum nodes are predicted to perform quantum tasks such as the classification of quantum entangled states, accessing of quantum state properties, and the preparation of quantum states.
See also
References
- ↑ Ghosh, Sanjib; Opala, Andrzej; Matuszewski, Michał; Paterek, Tomasz; Liew, Timothy C. H. (2019). "Quantum reservoir processing". NPJ Quantum Information. 5: 35. arXiv:1811.10335. Bibcode:2019npjQI...5...35G. doi:10.1038/s41534-019-0149-8.
This article "Quantum reservoir processor" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Quantum reservoir processor. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.
