Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2003.01605 · Physical Review Research

Neural-network approach for identifying nonclassicality from click-counting data

Valentin Gebhart, Martin Bohmann

Published May 11, 2020• 20 citations
arXivHTMLPDF

Abstract

This work uses the supervised learning of a deep neural network to identify nonclassicality of light detected through multiplexed measurement schemes. The network learns to correctly classify different classical and nonclassical states, outperforms other indicators of nonclassicality for a wide parameter range and also identifies nonclassicality for states outside the training set.

DOI
10.1103/physrevresearch.2.023150

Related papers

Learning to detect optical nonclassicality

Martina Jung, Suchitra Krishnaswamy, Timon Schapeler +4 more

Nonclassicality, defined in the quantum optical sense, serves as a resource for photon-based quantum technologies. Therefore, certifying the nonclassicality of a quantum state is crucial for gauging its potential for quantum advantage. However, traditional nonclassicality witnesses that assume perfect knowledge of the witness observables often fail in realistic scenarios with limited statistics and finite-resolution photon detectors. Furthermore, these witnesses do not exploit the fact that certain states are unlikely to be observed in a given experiment. Here, we train a variational model to distinguish classical from nonclassical states using finitely many measurement samples of multimode quantum states that are probed with different photon-number-resolving detection schemes. The learned decision rule is then an indicator of nonclassicality, tailored to a given set of physically relevant states. Our approach is both data-driven and interpretable in the sense that the learned analytical decision rule can be extracted. Training the model on experimental data measured with (i) a superconducting nanowire single-photon detector and (ii) a time-bin multiplexing detection scheme demonstrates the versatility of the approach, paving the way for efficient nonclassicality detection.

quant-pharXiv:2603.06319•Mar 9, 2026•1 citation

Identifying nonclassicality from experimental data using artificial neural networks

Valentin Gebhart, Martin Bohmann, Karsten Weiher +4 more

The authors use machine learning methods to identify nonclassicality from experimental data obtained via homodyne detection, opening the possibility for fast sorting and direct monitoring of quantum light in experiments.

arXiv:2101.07112•Physical Review Research•Jun 22, 2021•10 citations

Deep Neural Networks as the Semi-classical Limit of Topological Quantum Neural Networks: The problem of generalisation

Antonino Marcianò, Emanuele Zappala, Torda, Tommaso +5 more

Deep Neural Networks miss a principled model of their operation. A novel framework for supervised learning based on Topological Quantum Field Theory that looks particularly well suited for implementation on quantum processors has been recently explored. We propose using this framework to understand the problem of generalisation in Deep Neural Networks. More specifically, in this approach, Deep Neural Networks are viewed as the semi-classical limit of Topological Quantum Neural Networks. A framework of this kind explains the overfitting behavior of Deep Neural Networks during the training step and the corresponding generalisation capabilities. We explore the paradigmatic case of the perceptron, which we implement as the semiclassical limit of Topological Quantum Neural Networks. We apply a novel algorithm we developed, showing that it obtains similar results to standard neural networks, but without the need for training (optimisation).

arXiv:2210.13741•Oct 25, 2022•6 citations

Breast Cancer Detection Using Quantum Convolutional Neural Networks: A Demonstration on a Quantum Computer

Aradh Bisarya, Walid El Maouaki, Sabyasachi Mukhopadhyay +5 more

Deep learning have paved the way for scientists to achieve great technical feats. In an endeavor to hone and perfect these techniques, quantum deep learning is a promising and important tool to utilize to the fullest. Using the techniques of deep learning and supervised learning in a quantum framework, we are able to propose a quantum convolutional neural network and showcase its implementation. Using the techniques of deep learning and supervised learning in a quantum framework, we are able to propose a quantum convolutional neural network and showcase its implementation. We keep our focus on training of the ten qubits system in a way so that it can learn from labeling of the breast cell data of Wisconsin breast cancer database and optimize the circuit parameters to obtain the minimum error. Through our study, we also have showcased that quantum convolutional neural network can outperform its classical counterpart not only in terms of accuracy but also in the aspect of better time complexity.

•Jun 23, 2020•30 citations

Supervised machine learning of ultracold atoms with speckle disorder

Sebastiano Pilati, P. Pieri

We analyze how accurately supervised machine learning techniques can predict the lowest energy levels of one-dimensional noninteracting ultracold atoms subject to the correlated disorder due to an optical speckle field. Deep neural networks with different numbers of hidden layers and neurons per layer are trained on large sets of instances of the speckle field, whose energy levels have been preventively determined via a high-order finite difference technique. The Fourier components of the speckle field are used as the feature vector to represent the speckle-field instances. A comprehensive analysis of the details that determine the possible success of supervised machine learning tasks, namely the depth and the width of the neural network, the size of the training set, and the magnitude of the regularization parameter, is presented. It is found that ground state energies of previously unseen instances can be predicted with an essentially negligible error given a computationally feasible number of training instances. First and second excited state energies can be predicted too, albeit with slightly lower accuracy and using more layers of hidden neurons. We also find that a three-layer neural network is remarkably resilient to Gaussian noise added to the training-set data (up to 10% noise level), suggesting that cold-atom quantum simulators could be used to train artificial neural networks.

arXiv:1809.04679•Scientific Reports•Apr 4, 2019•39 citations

Revealing nonclassicality of multiphoton optical beams via artificial neural networks

Radek Machulka, Jan Peřina, Václav Michálek +2 more

The identification of nonclassical features of multiphoton quantum states represents a crucial task in the development of many quantum photonic technologies. Under realistic experimental conditions, a photonic quantum state gets affected by its interaction with several nonideal optoelectronic devices, including those used to guide, detect, or characterize it. The result of such noisy interaction is that the nonclassical features of the original quantum state get considerably reduced or are completely absent in the detected, final state. In this work, the self-learning features of artificial neural networks are exploited to experimentally show that the nonclassicality of multiphoton quantum states can be assessed and fully characterized, even in the cases in which the nonclassical features are concealed by the measuring devices. Our work paves the way toward artificial-intelligence-assisted experimental-setup characterization, as well as quantum state nonclassicality identification. Published by the American Physical Society 2024

arXiv:2405.09174•Physical Review Applied•Sep 20, 2024•0 citations
This work uses the supervised learning of a deep neural network to identify nonclassicality of light detected through multiplexed measurement schemes. The network learns to correctly classify different classical and nonclassical states, outperforms other indicators of nonclassicality for a wide parameter range and also identifies nonclassicality for states outside the training set.
Nonclassicality, defined in the quantum optical sense, serves as a resource for photon-based quantum technologies. Therefore, certifying the nonclassicality of a quantum state is crucial for gauging its potential for quantum advantage. However, traditional nonclassicality witnesses that assume perfect knowledge of the witness observables often fail in realistic scenarios with limited statistics and finite-resolution photon detectors. Furthermore, these witnesses do not exploit the fact that certain states are unlikely to be observed in a given experiment. Here, we train a variational model to distinguish classical from nonclassical states using finitely many measurement samples of multimode quantum states that are probed with different photon-number-resolving detection schemes. The learned decision rule is then an indicator of nonclassicality, tailored to a given set of physically relevant states. Our approach is both data-driven and interpretable in the sense that the learned analytical decision rule can be extracted. Training the model on experimental data measured with (i) a superconducting nanowire single-photon detector and (ii) a time-bin multiplexing detection scheme demonstrates the versatility of the approach, paving the way for efficient nonclassicality detection.
Deep Neural Networks miss a principled model of their operation. A novel framework for supervised learning based on Topological Quantum Field Theory that looks particularly well suited for implementation on quantum processors has been recently explored. We propose using this framework to understand the problem of generalisation in Deep Neural Networks. More specifically, in this approach, Deep Neural Networks are viewed as the semi-classical limit of Topological Quantum Neural Networks. A framework of this kind explains the overfitting behavior of Deep Neural Networks during the training step and the corresponding generalisation capabilities. We explore the paradigmatic case of the perceptron, which we implement as the semiclassical limit of Topological Quantum Neural Networks. We apply a novel algorithm we developed, showing that it obtains similar results to standard neural networks, but without the need for training (optimisation).
Deep learning have paved the way for scientists to achieve great technical feats. In an endeavor to hone and perfect these techniques, quantum deep learning is a promising and important tool to utilize to the fullest. Using the techniques of deep learning and supervised learning in a quantum framework, we are able to propose a quantum convolutional neural network and showcase its implementation. Using the techniques of deep learning and supervised learning in a quantum framework, we are able to propose a quantum convolutional neural network and showcase its implementation. We keep our focus on training of the ten qubits system in a way so that it can learn from labeling of the breast cell data of Wisconsin breast cancer database and optimize the circuit parameters to obtain the minimum error. Through our study, we also have showcased that quantum convolutional neural network can outperform its classical counterpart not only in terms of accuracy but also in the aspect of better time complexity.
We analyze how accurately supervised machine learning techniques can predict the lowest energy levels of one-dimensional noninteracting ultracold atoms subject to the correlated disorder due to an optical speckle field. Deep neural networks with different numbers of hidden layers and neurons per layer are trained on large sets of instances of the speckle field, whose energy levels have been preventively determined via a high-order finite difference technique. The Fourier components of the speckle field are used as the feature vector to represent the speckle-field instances. A comprehensive analysis of the details that determine the possible success of supervised machine learning tasks, namely the depth and the width of the neural network, the size of the training set, and the magnitude of the regularization parameter, is presented. It is found that ground state energies of previously unseen instances can be predicted with an essentially negligible error given a computationally feasible number of training instances. First and second excited state energies can be predicted too, albeit with slightly lower accuracy and using more layers of hidden neurons. We also find that a three-layer neural network is remarkably resilient to Gaussian noise added to the training-set data (up to 10% noise level), suggesting that cold-atom quantum simulators could be used to train artificial neural networks.
The identification of nonclassical features of multiphoton quantum states represents a crucial task in the development of many quantum photonic technologies. Under realistic experimental conditions, a photonic quantum state gets affected by its interaction with several nonideal optoelectronic devices, including those used to guide, detect, or characterize it. The result of such noisy interaction is that the nonclassical features of the original quantum state get considerably reduced or are completely absent in the detected, final state. In this work, the self-learning features of artificial neural networks are exploited to experimentally show that the nonclassicality of multiphoton quantum states can be assessed and fully characterized, even in the cases in which the nonclassical features are concealed by the measuring devices. Our work paves the way toward artificial-intelligence-assisted experimental-setup characterization, as well as quantum state nonclassicality identification. Published by the American Physical Society 2024