Abstract
We demonstrate that neural networks that process noisy data can learn to exploit, when available, access to auxiliary noise that is correlated with the noise on the data. In effect, the network learns to use the correlated auxiliary noise as an approximate key to decipher its noisy input data. An example of naturally occurring correlated auxiliary noise is the noise due to decoherence. Our results could, therefore, also be of interest, for example, for machine-learned quantum error correction.
Highlights
We demonstrate that neural networks that process noisy data can learn to exploit, when available, access to auxiliary noise that is correlated with the noise on the data
Our first aim is to show that neural networks that learn a task on noisy data, such as, e.g., image classification, can simultaneously learn to improve their performance by exploiting access to separate noise that is correlated with the noise in the data, when such auxiliary correlated noise is available
The novel UCAN approach is, not primarily concerned with traditional denoising, see, e.g.,4–11, but is instead concerned with new opportunities for neural networks that arise in the event of the availability of correlated auxiliary noise
Summary
We demonstrate that neural networks that process noisy data can learn to exploit, when available, access to auxiliary noise that is correlated with the noise on the data. The network learns to improve its performance by implicitly using the auxiliary correlated noise to subtract some of the noise from the data This new approach of ‘Utilizing Correlated Auxiliary Noise’ (UCAN), has potential applications, for example, whenever noise arising in a measurement is correlated with noise that can be picked up in a vicinity of the measurement. There, one application could be to the main bottleneck for quantum computing technology, the process of decoherence This is because decoherence consists of the generating of correlated auxiliary noise in degrees of freedom in the immediate environment of the physical qubits. This could yield a novel form of machine-learned quantum error correction that is not based on traditional quantum error-correction principles such as utilizing redundant coding or topologiocal stability but that instead tries to access environmental degrees of freedom to re-integrate previously leaked quantum information into the circuit
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
