IFISC (UIB-CSIC) research on Quantum Reservoir Computing is highlighted by the editorial staff of Optics Express

Feb. 27, 2024

  • An IFISC study, published in Optics Express and featured as an Editor's Pick, reveals the potential of quantum squeezing in the machine learning architecture known as Reservoir Computing.

A recent article published by IFISC researchers in Optics Express, entitled "Squeezing as a resource for time series processing in quantum reservoir computing", has been selected by the journal's editorial team as Editor's Pick, highlighting its relevance and excellence in the scientific field. This distinction recognises research papers that not only demonstrate exceptional scientific quality, but also represent significant contributions to their respective areas of study.

The study analyses a novel architecture for Quantum Reservoir Computing (QRC), based on a photonic loop in which light (multimode) flows in a cavity with a non-linear medium. At each time step, the input is encoded in the signal and the fluctuations of the light are measured by extracting a part of the beam. This platform exploits the phenomenon of quantum squeezing, which is a reduction of the quantum fluctuations of light below the level of shot noise. This is used to optimise time series processing, a key challenge in computing and data analysis. Quantum squeezing is explored by researchers as a tool to improve the memory capacity and performance of the quantum reservoir.

Jorge García-Beni, Gian Luca Giorgi, Miguel C. Soriano and Roberta Zambrini, authors of the paper, demonstrate how this quantum squeezing can significantly increase the accessible memory of the reservoir, resulting in a remarkable performance improvement in various temporal data processing tasks in terms of both recall and prediction. The results reveal that, depending on the model and experimental noise, quantum compression can play both a beneficial and detrimental role in QRC performance, so characterising the system is key. This finding underlines the importance of a thorough understanding of quantum dynamics and its interaction with the experimental environment to optimally design new quantum computing systems.

These results advance the understanding of the role of concepts such as entanglement and quantum compression in neuromorphic learning and information processing. Quantum compression is revealed as a key quantum resource for advanced applications in metrology, cryptography and computation.

This recognition by the journal as an Editor's Pick underlines the importance of the authors' work in advancing quantum computing and its application in time series processing. The study marks a breakthrough in exploring the limits of quantum technology to solve complex problems and improve our information processing capabilities.

Jorge García-Beni, Gian Luca Giorgi, Miguel C. Soriano, and Roberta Zambrini, "Squeezing as a resource for time series processing in quantum reservoir computing", Opt. Express 32, 6733-6747 (2024). DOI: https://doi.org/10.1364/OE.507684

Image: Time series prediction for the Mackey-Glass system, where the actual attractor (white line and dots) is compared with the model predictions (green line and dots).


Related Research projects


Quantum machine learning using reservoir computing

P.I.: Miguel C. Soriano, Roberta Zambrini
The QuaResC project engages in a new collaboration UIB and CSIC researchers at IFISC with the objective to address an interdisciplinary topic between artificial intelligence and quantum physics: quantum machine learning using ...

This web uses cookies for data collection with a statistical purpose. If you continue browsing, it means acceptance of the installation of the same.

More info I agree