Quantum Machine Learning: An Interplay Between Quantum Computing and Machine Learning

Chao-Han Yang; Jun Qi; Pin-Yu Chen; Samuel Yen-Chi Chen

REVIEW 4 major objections 4 minor 44 references

Quantum Machine Learning: An Interplay Between Quantum Computing and Machine Learning

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This review argues that variational quantum circuits are a viable NISQ-era machine-learning architecture, reporting better speech recognition with QCNN features and a BERT-QCNN hybrid that surpasses classical deep learning methods.

desk verdict A readable recap of the authors' own QML line, with no new results; the empirical claims are unverifiable in the text and partly contradicted by the paper's own limitations section. read the letter →

arxiv 2411.09403 v1 pith:AX6H67SQ submitted 2024-11-14 quant-ph cs.AI

Jun Qi , Chao-Han Yang , Samuel Yen-Chi Chen , Pin-Yu Chen This is my paper · ORCID

classification quant-phcs.AI

keywords quantummachinelearningvariationalcircuitsconvolutionalneuralnetworkshybridquantum-classicalNISQdevicescircuitarchitecturesearchtensor-trainspeechrecognition

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This review-style paper argues that variational quantum circuits (VQCs), parameterized quantum circuits trained like neural networks, are a workable architecture for quantum machine learning on noisy intermediate-scale quantum (NISQ) devices. The authors report that quantum convolutional neural networks (QCNNs) extract speech features that are more discriminative than classical CNN features and yield better spoken-language understanding accuracy, and that a hybrid BERT-QCNN model surpasses leading classical deep learning methods in text classification. They also argue that classical machine learning tools, including tensor-train networks, pre-trained models, and generative or reinforcement-learning-based circuit search, can strengthen VQC representation and automate circuit design. If the claims hold, hybrid quantum-classical models could deliver practical accuracy gains before fault-tolerant quantum computers exist. The paper grounds these claims in classical simulations and explicitly assumes the existence of quantum logic qubits.

What carries the argument

The load-bearing mechanism is the variational quantum circuit (VQC): a small quantum circuit that encodes classical data through Pauli $R_Y$ rotations, applies entangling CNOT gates together with tunable $R_X$, $R_Y$, and $R_Z$ rotations, and reads out expectation values of Pauli-$Z$ observables after repeated measurement. The quantum convolutional neural network (QCNN) uses such VQC blocks in place of classical convolution filters to extract features. The hybrid architectures add classical machinery around the VQC: a tensor-train network (TTN) reduces input dimensionality before the VQC, and a frozen BERT embedding feeds text tokens into a QCNN whose only trained parameters are the circuit angles. In the paper's account, this division of labor is what makes VQCs trainable and lets classical optimization methods, such as stochastic gradient descent, adjust the circuit parameters.

What would settle it

Run the QCNN speech-recognition and BERT-QCNN text-classification pipelines on a real noisy near-term quantum processor, with the same preprocessing, optimizer, and baselines, and compare accuracies. If the gains over classical CNN and BERT baselines disappear or the circuits fail to train at realistic error rates, the paper's central claim that VQCs improve real-world machine learning tasks would be refuted.

Watch

Extended reading notes

Core claim

The central claim is that variational quantum circuits constitute a useful QML architecture and that combining them with classical models extends their reach on NISQ hardware. On the authors' terms, the discovery is empirical: quantum convolution maps Mel-spectrogram speech features into a representation that is more discriminative than the original spectrogram or CNN-encoded features, and this leads to even better speech recognition accuracy; a BERT-QCNN text classifier in which BERT's parameters stay frozen while VQC parameters are fine-tuned surpasses the performance of leading classical deep learning methods. A complementary theoretical result is the approximation bound $\mathcal{O}(1/\sqrt{U})+\mathcal{O}(1/\sqrt{M})$ for VQC-based functional regression, which the authors use to explain why adding a tensor-train front end (TTN-VQC) improves representation power under limited qubit counts. The whole picture is presented as an architecture plus a set of results, not as a single theorem.

Load-bearing premise

The central claim holds only if trainable quantum circuits can run accurately enough on real near-term hardware; the paper admits that its research mostly relies on classical simulations that assume working quantum qubits.

Editorial extensions

If this is right

Quantum convolution can serve as a feature extraction front end for automatic speech recognition, producing representations that are more discriminative than classical CNN features.
Pre-trained classical language models can be paired with quantum circuits in frozen-backbone mode, so quantum fine-tuning needs only the circuit parameters to be updated.
The $\mathcal{O}(1/\sqrt{U})+\mathcal{O}(1/\sqrt{M})$ approximation bound means representation error improves slowly as qubits and measurements increase, making tensor-train or other classical pre-processing a practical necessity at NISQ scale.
Generative models and reinforcement learning can automate quantum circuit architecture search, which may reduce the optimization difficulties of training deep VQCs.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the reported gains come from classical simulation, the more direct test is to run the same pipelines on hardware and measure accuracy as a function of two-qubit gate error rate; that curve would show how much noise the QCNN advantage tolerates.
The paper does not isolate the quantum contribution; replacing the QCNN block in BERT-QCNN with a classical layer of matched parameter count would reveal whether entanglement or just an extra trainable layer drives the text-classification gains.
If the mechanism is general, the same VQC feature extractor should transfer to image or sensor data; a positive result there would strengthen the case that quantum convolution, rather than a speech-specific artifact, is doing the work.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: this is not a research paper; it's a survey of the authors' prior work on VQC-based QML. There is no new architecture, theorem, dataset, or experiment. The only equations are standard definitions and the Q-learning loss from their earlier papers. So if you're looking for a new result, skip it.

What it does well: it organizes a sprawling line of work into a coherent narrative. The VQC setup, quantum RL, QCNN for speech, TTN-VQC error bound, BERT-QCNN, and quantum architecture search are all laid out with figures and references. Someone new to this group's research could use it as a map. The reference list is broad and the text is readable.

Soft spots, in order of importance. First, the central performance claims are made in passing with no evidence: QCNN features 'result in even better speech recognition accuracy,' and BERT-QCNN 'surpasses the performance of leading classical deep learning methods,' but there are no datasets, metrics, baselines, qubit counts, or error bars anywhere. You cannot check these statements from the manuscript. Second, the paper's own limitations section says the research 'primarily relies on classical simulations, assuming the existence of quantum logic qubits,' which sits awkwardly with the earlier assertion that VQCs are 'resilient to the quantum noise on NISQ devices.' The cited references for that claim, [17,18], are about generalization and feature Hilbert spaces; their titles don't support noise resilience. Third, the BERT-QCNN comparison is confounded: BERT's parameters are fixed, so any gain could come from the classical side. These are transparency issues, not necessarily fatal to the underlying research, but they make the manuscript's central claim unverified.

The citation pattern is heavily self-referential, which is expected in a summary of one's own work. It's not a flaw by itself, but combined with the unverifiable claims, it means the paper cannot stand as evidence for anything beyond 'this is what we have published.'

Who is this for? A reader who wants a short overview of this group's QML portfolio, or a course reading to see how VQCs are applied across tasks. Not for someone hunting for new technical contributions.

Recommendation: I would not send this to peer review as a research article. If it were submitted as a review/survey, it would need major revisions: either add real experiment summaries with numbers or temper every performance claim to match what the citations actually show. As it stands, it's a useful but non-archival memo.

Referee Report

4 major / 4 minor

Summary. The paper is a short position/summary article in which the authors describe their recent work on quantum machine learning (QML). It introduces variational quantum circuits (VQC) as a QML architecture, discusses quantum reinforcement learning and quantum convolutional neural networks (QCNN), and then presents hybrid quantum-classical approaches such as TTN-VQC and BERT-QCNN, together with quantum circuit architecture search. The main claims are that VQC-based models are resilient to NISQ noise and that the authors' QCNN-based speech features and BERT-QCNN text classifier achieve superior real-world accuracy. However, the paper contains no experimental data, derivations, or comparisons; it relies on qualitative assertions and citations to the authors' previous papers, and it later concedes that the research primarily uses classical simulations assuming logical qubits.

Significance. If the central performance claims were established, the paper would point to a practically useful role for VQC-based hybrid models on near-term hardware. The strongest parts of the manuscript are its concise presentation of the VQC formalism (Sec. II-A), the clear figures illustrating quantum circuits and hybrid architectures, and the honest acknowledgment in Sec. IV that the work is based on classical simulations. The paper also usefully surveys the authors' own prior publications and connects them to broader QML questions, including quantum circuit architecture search. That said, the paper is not self-contained: its headline claims of improved speech recognition accuracy and of surpassing classical deep learning methods are stated without any numerical evidence, baselines, or error bars, so the significance for the broader community cannot be assessed from this manuscript alone.

major comments (4)

[Sec. II-C] The claim that QCNN-encoded speech features 'result in even better speech recognition accuracy' is load-bearing for the paper's central thesis, yet the manuscript provides no dataset, metric, baseline, number of qubits, or error bars to support it. The sentence immediately preceding it says 'in our experiments of spoken language understanding,' but no experimental details appear anywhere in the paper, so a reader cannot verify or reproduce the result.
[Sec. III-A] The statement that 'Our classical simulations on CPU/GPU and real-world quantum experiments demonstrate that the BERT-QCNN model surpasses the performance of leading classical deep learning methods' is another central claim, but the paper reports no comparison methods, no accuracy numbers, no experimental setup, and no statistical significance. In addition, because BERT's parameters are fixed in this hybrid architecture, any observed improvement could plausibly come from BERT's classical representation rather than from the quantum component, so the claim of a quantum advantage is not established.
[Sec. IV] The paper's own limitation statement, 'our research primarily relies on classical simulations, assuming the existence of quantum logic qubits,' directly contradicts the earlier assertion in Sec. I that VQCs 'have been demonstrated to be resilient to the quantum noise on NISQ devices.' The manuscript must either reconcile these statements or explicitly qualify the hardware claims, because the NISQ-fidelity premise is essential to the paper's argument that VQC-based QML can improve real-world tasks on current hardware.
[Sec. I] The statement that VQCs 'have been demonstrated to be resilient to the quantum noise on NISQ devices [17], [18]' is not supported by the cited references: [17] is a theoretical work on generalization bounds from few training data, and [18] is a theoretical work on feature Hilbert spaces. Neither citation demonstrates empirical noise resilience on NISQ hardware. This citation either needs to be corrected or the claim needs to be removed or reworded.

minor comments (4)

[Author affiliation] The affiliation line contains a typo: 'Hong Kong Baptist Univeristy' should be 'Hong Kong Baptist University.'
[Sec. II-C] In the sentence 'Figure 4 compares the signal signal features encoded by classical CNN and QCNN models,' the word 'signal' is repeated; this should be corrected.
[Sec. III-A heading] The heading 'Hybrid Quantum-Classical Neural Neworks' contains a typo; 'Neworks' should be 'Networks.'
[Sec. II-A, Eq. (1)] The notation uses U both as the number of qubits and as the index upper bound in the tensor product; this is not wrong, but it would be clearer to use a different symbol for the upper bound, such as n, to avoid confusion with the unitary gates denoted U.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the paper is a research summary whose claims rest on prior peer-reviewed work, not on definitions or fitted parameters that reproduce the target results.

full rationale

This manuscript is an overview of the authors' prior QML results; it does not present a derivation chain in which an output quantity is defined in terms of the quantity it purports to predict. Section II-A defines the standard VQC encoding and parametric circuit using elementary gates, and Eq. (2) is the conventional Q-learning loss; none of these define a target result in terms of itself. The empirical statements (QCNN speech features 'result in even better speech recognition accuracy'; BERT-QCNN 'surpasses the performance of leading classical deep learning methods') are qualitative attributions to the authors' earlier peer-reviewed papers [26] and [35], not predictions computed from inputs in this manuscript. No dataset, metric, fitted parameter, or numerical result appears in the text, so there is no 'fitted input called prediction' and no equation that reduces to another equation by construction. The TTN-VQC error bound is cited as a theorem from [32] rather than derived here, and the noise-resilience claim is supported by [17,18] that do not obviously demonstrate it. Section IV concedes that 'our research primarily relies on classical simulations, assuming the existence of quantum logic qubits,' which is a serious evidence/scope limitation for the NISQ framing but is not a circularity. Although self-citations are abundant, they point to externally published, potentially checkable work, and under the stated rules such citations do not by themselves raise the circularity score. No specific circular step can be exhibited.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central claim rests on five unproved background assumptions: a universal approximation theorem for quantum gates, the availability of usable quantum qubits, the accuracy of the cited speed-up table, the noise resilience of VQCs, and a cited error bound. None of these is established inside this paper. The only fitted quantities mentioned are the VQC angles, whose trained values are not reported. The evidence for all performance statements is external citation to the authors' own prior work.

free parameters (1)

Variational circuit rotation angles (alpha_i, beta_i, gamma_i) = not reported in this manuscript
The performance claims depend on trained VQC parameters, but no values, training setups, or final accuracies are given here; they are inherited from cited prior papers.

assumptions (5)

standard math Universal approximation theorem for quantum circuits: any continuous function can be approximated using one- and two-qubit gates.
Invoked in Section II without proof or citation; used to justify VQC expressivity.
domain assumption Existence of usable quantum logic qubits on NISQ devices for training VQCs.
Section IV concedes the research 'primarily relies on classical simulations, assuming the existence of quantum logic qubits'; without this, the hardware claims do not apply.
domain assumption The speed-up table in Table I accurately reflects known quantum algorithms.
The table is copied from reference [7] and used to motivate QML, but no caveats or conditions are discussed in this paper.
domain assumption VQC is resilient to quantum noise on NISQ devices.
Stated in Section I with support only from cited references [17], [18]; no noise experiment appears in this document.
domain assumption The error bound O(1/sqrt(U)) + O(1/sqrt(M)) for the VQC functional regression is correct.
Stated in Section III-A as a known result from reference [32], with no derivation or proof included.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantum Machine Learning: An Interplay Between Quantum Computing and Machine Learning." pith.science (2026). https://pith.science/paper/AX6H67SQ

@misc{pith2026241109403,
  author       = {Pith},
  title        = {Pith review of: Quantum Machine Learning: An Interplay Between Quantum Computing and Machine Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AX6H67SQ}},
  note         = {Machine review of arXiv:2411.09403}
}

read the original abstract

Quantum machine learning (QML) is a rapidly growing field that combines quantum computing principles with traditional machine learning. It seeks to revolutionize machine learning by harnessing the unique capabilities of quantum mechanics and employs machine learning techniques to advance quantum computing research. This paper introduces quantum computing for the machine learning paradigm, where variational quantum circuits (VQC) are used to develop QML architectures on noisy intermediate-scale quantum (NISQ) devices. We discuss machine learning for the quantum computing paradigm, showcasing our recent theoretical and empirical findings. In particular, we delve into future directions for studying QML, exploring the potential industrial impacts of QML research.

Figures

Figures reproduced from arXiv: 2411.09403 by the authors.

**Figure 1.** Illustration of Variational Quantum Circuits. structure. The PQC model in the green dashed square is repeatedly copied to build a deep PQC architecture, which outputs U quantum states |o1⟩, |o2⟩, ..., |oU ⟩. The measurement transforms the quantum states |oi⟩ into the expected values ⟨σi⟩ = ⟨oi |σ (i) z |oi⟩ associated with the observables of PauliZ matrices σ (i) z . We use an arithmetic average of M times’ quantum… view at source ↗

**Figure 2.** Illustration of Quantum Reinforcement Learning. Quantum Q-learning stands for the main approach to QRL. Quantum Q-learning, using a VQC as the agent, learns the best possible action-value function without following a specific policy. It starts with a random initialized Qπ (s, a) for all states s ∈ S and actions a ∈ A, and it uses the VQC to represent Qπ (s, a). The experience replay technique is employed to store pa… view at source ↗

**Figure 5.** An illustration of TTN-VQC structure. theory to scale up the QML for more complicated tasks. For one thing, we exploit a hybrid quantum-classical neural network to improve quantum models’ representation and generalization powers. For another, we rely on the generative AI for quantum circuit architecture search. A. Hybrid Quantum-Classical Neural Neworks Hybrid quantum-classical neural networks combine the power of … view at source ↗

Figures from the paper (2 more)

**Figure 6.** Figure 6: Hybrid BERT-QCNN model for text classification. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p003_6.png]

**Figure 7.** Figure 7: Using generative models for quantum circuit architecture search. and QCNN for text classification [35]. BERT [36], a widelyused pre-trained language model, enhances the QCNN’s quantum circuits’ ability to represent text data. BERT’s parameters remain unchanged in thi…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

44 extracted references · 37 canonical work pages

[17]

Generalization In Quantum Machine Learning From Few Training Data,

Matthias C Caro et al., “Generalization In Quantum Machine Learning From Few Training Data,” Nature Communications, vol. 13, no. 1, pp. 4919, 2022

work page 2022
[18]

Quantum Machine Learning in Feature Hilbert Spaces,

Maria Schuld and Nathan Killoran, “Quantum Machine Learning in Feature Hilbert Spaces,” Physical Review Letters , vol. 122, no. 4, pp. 040504, 2019

work page 2019
[1]

Li Deng, Deep Learning in Natural Language Processing , Springer, 2018

work page 2018
[2]

Richard Szeliski, Computer Vision: Algorithms and Applications , Springer Nature, 2022

work page 2022
[3]

AI-Powered Drug Discovery Captures Pharma Interest,

Eric Smalley, “AI-Powered Drug Discovery Captures Pharma Interest,” Nature Biotechnology, vol. 35, no. 7, pp. 604–606, 2017

work page 2017
[4]

Michael A Nielsen and Isaac L Chuang, Quantum Computation and Quantum Information, Cambridge University Press, 2010

work page 2010
[5]

Fault-Tolerant Logical Gates in The IBM Quantum Experience,

Robin Harper and Steven T Flammia, “Fault-Tolerant Logical Gates in The IBM Quantum Experience,” Physical review letters , vol. 122, no. 8, pp. 080504, 2019

work page 2019
[6]

CUDA Quantum: The Platform for Integrated Quantum-Classical Computing,

Jin-Sung Kim, Alex McCaskey, Bettina Heim, Manish Modani, Sam Stanwyck, and Timothy Costa, “CUDA Quantum: The Platform for Integrated Quantum-Classical Computing,” in 2023 60th ACM/IEEE Design Automation Conference . IEEE, 2023, pp. 1–4

work page 2023

Show all 44 references

[7]

Quantum Machine Learning,

Jacob Biamonte, Peter Wittek, Nicola Pancotti, Patrick Rebentrost, Nathan Wiebe, and Seth Lloyd, “Quantum Machine Learning,” Nature, vol. 549, no. 7671, pp. 195–202, 2017

2017
[8]

Power of Data in Quantum Machine Learning,

Hsin-Yuan Huang et al., “Power of Data in Quantum Machine Learning,” Nature Communications, vol. 12, no. 1, pp. 1–9, 2021

2021
[9]

An Intro- duction to Quantum Machine Learning,

Maria Schuld, Ilya Sinayskiy, and Francesco Petruccione, “An Intro- duction to Quantum Machine Learning,” Contemporary Physics , vol. 56, no. 2, pp. 172–185, 2015

2015
[10]

A Rigor- ous and Robust Quantum Speed-up in Supervised Machine Learning,

Yunchao Liu, Srinivasan Arunachalam, and Kristan Temme, “A Rigor- ous and Robust Quantum Speed-up in Supervised Machine Learning,” Nature Physics, vol. 17, no. 9, pp. 1013–1017, 2021

2021
[11]

Towards Provably Efficient Quantum Algorithms for Large-scale Machine-Learning Models,

Junyu Liu et al., “Towards Provably Efficient Quantum Algorithms for Large-scale Machine-Learning Models,” Nature Communications, vol. 15, no. 1, pp. 434, 2024

2024
[12]

Quantum Computing in the NISQ Era and Beyond,

John Preskill, “Quantum Computing in the NISQ Era and Beyond,” Quantum, vol. 2, pp. 79, August 2018

2018
[13]

Challenges and Opportunities in Quantum Machine Learning,

M Cerezo, Guillaume Verdon, Hsin-Yuan Huang, Lukasz Cincio, and Patrick J Coles, “Challenges and Opportunities in Quantum Machine Learning,” Nature Computational Science , vol. 2, no. 9, pp. 567–576, 2022

2022
[14]

Variational Quantum Algorithms,

Marco Cerezo et al., “Variational Quantum Algorithms,” Nature Reviews Physics, vol. 3, no. 9, pp. 625–644, 2021

2021
[15]

Quantum Convolu- tional Neural Networks,

Iris Cong, Soonwon Choi, and Mikhail D Lukin, “Quantum Convolu- tional Neural Networks,” Nature Physics, vol. 15, no. 12, pp. 1273–1278, 2019

2019
[16]

Quantum Graph Neural Net- works,

Guillaume Verdon, Trevor McCourt, Enxhell Luzhnica, Vikash Singh, Stefan Leichenauer, and Jack Hidary, “Quantum Graph Neural Net- works,” arXiv preprint arXiv:1909.12264 , 2019

1909 arXiv
[19]

Variational Quantum Circuits for Deep Reinforcement Learning,

Samuel Yen-Chi Chen et al., “Variational Quantum Circuits for Deep Reinforcement Learning,” IEEE Access , vol. 8, pp. 141007–141024, 2020

2020
[20]

Variational quantum reinforcement learning via evolutionary optimization,

Samuel Yen-Chi Chen, Chih-Min Huang, Chia-Wei Hsing, Hsi-Sheng Goan, and Ying-Jer Kao, “Variational quantum reinforcement learning via evolutionary optimization,” Machine Learning: Science and Tech- nology, vol. 3, no. 1, pp. 015025, 2022

2022
[21]

Quantum deep recurrent reinforcement learn- ing,

Samuel Yen-Chi Chen, “Quantum deep recurrent reinforcement learn- ing,” in ICASSP 2023-2023 IEEE International Conference on Acous- tics, Speech and Signal Processing (ICASSP) . IEEE, 2023, pp. 1–5

2023
[22]

Asynchronous training of quantum reinforce- ment learning,

Samuel Yen-Chi Chen, “Asynchronous training of quantum reinforce- ment learning,” Procedia Computer Science , vol. 222, pp. 321–330, 2023

2023
[23]

Quantum deep Q-learning with distributed prioritized experience replay,

Samuel Yen-Chi Chen, “Quantum deep Q-learning with distributed prioritized experience replay,” in 2023 IEEE International Conference on Quantum Computing and Engineering. IEEE, 2023, vol. 2, pp. 31–35

2023
[24]

On Mean Absolute Error for Deep Neural Network Based Vector-to- Vector Regression,

Jun Qi, Jun Du, Sabato Marco Siniscalchi, Xiaoli Ma, and Chin-Hui Lee, “On Mean Absolute Error for Deep Neural Network Based Vector-to- Vector Regression,” IEEE Signal Processing Letters , vol. 27, pp. 1485– 1489, 2020

2020
[25]

Quantum convolutional neural networks for high energy physics data analysis,

Samuel Yen-Chi Chen, Tzu-Chieh Wei, Chao Zhang, Haiwang Yu, and Shinjae Yoo, “Quantum convolutional neural networks for high energy physics data analysis,” Physical Review Research , vol. 4, no. 1, pp. 013231, 2022

2022
[26]

Decen- tralizing feature extraction with quantum convolutional neural network for automatic speech recognition,

Chao-Han Huck Yang, Jun Qi, Samuel Yen-Chi Chen, Pin-Yu Chen, Sabato Marco Siniscalchi, Xiaoli Ma, and Chin-Hui Lee, “Decen- tralizing feature extraction with quantum convolutional neural network for automatic speech recognition,” in ICASSP 2021-2021 IEEE In- ternational Confe...

2021
[27]

An End-to-End Trainable Hybrid Classical-Quantum Classifier,

Samuel Yen-Chi Chen, Chih-Min Huang, Chia-Wei Hsing, and Ying-Jer Kao, “An End-to-End Trainable Hybrid Classical-Quantum Classifier,” Machine Learning: Science and Technology , vol. 2, no. 4, pp. 045021, 2021

2021
[28]

QTN-VQC: An End-to-End Learning Framework for Quantum Neural Networks,

Jun Qi, Chao-Han Yang, and Pin-Yu Chen, “QTN-VQC: An End-to-End Learning Framework for Quantum Neural Networks,” Physica Scripta, vol. 99, no. 1, pp. 015111, 2023

2023
[29]

Tensor-Train Decomposition,

Ivan V Oseledets, “Tensor-Train Decomposition,” SIAM Journal on Scientific Computing, vol. 33, no. 5, pp. 2295–2317, 2011

2011
[30]

Exploiting Low-Rank Tensor-Train Deep Neural Networks Based on Riemannian Gradient Descent With Illustrations of Speech Processing,

Jun Qi, Chao-Han Huck Yang, Pin-Yu Chen, and Javier Tejedor, “Exploiting Low-Rank Tensor-Train Deep Neural Networks Based on Riemannian Gradient Descent With Illustrations of Speech Processing,” arXiv preprint arXiv:2203.06031 , 2022

2022 arXiv
[31]

Mitigating Clip- ping Distortion in Multicarrier Transmissions Using Tensor-Train Deep Neural Networks,

Muhammad Shahmeer Omar, Jun Qi, and Xiaoli Ma, “Mitigating Clip- ping Distortion in Multicarrier Transmissions Using Tensor-Train Deep Neural Networks,” IEEE Transactions on Wireless Communications, vol. 22, no. 3, pp. 2127–2138, 2022

2022
[32]

The- oretical Error Performance Analysis for Variational Quantum Circuit Based Functional Regression,

Jun Qi, Chao-Han Huck Yang, Pin-Yu Chen, and Min-Hsiu Hsieh, “The- oretical Error Performance Analysis for Variational Quantum Circuit Based Functional Regression,” npj Quantum Information , vol. 9, no. 1, pp. 4, 2023

2023
[33]

Pre-training Tensor-Train Networks Facilitates Machine Learning with Variational Quantum Circuits,

Jun Qi, Chao-Han Huck Yang, Pin-Yu Chen, and Min-Hsiu Hsieh, “Pre-training Tensor-Train Networks Facilitates Machine Learning with Variational Quantum Circuits,” arXiv preprint arXiv:2306.03741, 2023

2023
[34]

Classical-to-Quantum Transfer Learning for Spoken Command Recognition Based on Quantum Neural Networks,

Jun Qi and Javier Tejedor, “Classical-to-Quantum Transfer Learning for Spoken Command Recognition Based on Quantum Neural Networks,” in IEEE International Conference on Acoustics, Speech and Signal Processing, 2022, pp. 8627–8631

2022
[35]

When BERT Meets Quantum Temporal Convolution Learning for Text Classification in Heterogeneous Computing,

Chao-Han Huck Yang, Jun Qi, Samuel Yen-Chi Chen, Yu Tsao, and Pin-Yu Chen, “When BERT Meets Quantum Temporal Convolution Learning for Text Classification in Heterogeneous Computing,” in IEEE International Conference on Acoustics, Speech and Signal Processing , 2022

2022
[36]

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” in Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technolo- gies, 20...

2019
[37]

Quantum Circuit Architecture Search for Variational Quantum Algorithms,

Yuxuan Du, Tao Huang, Shan You, Min-Hsiu Hsieh, and Dacheng Tao, “Quantum Circuit Architecture Search for Variational Quantum Algorithms,” npj Quantum Information , vol. 8, no. 1, pp. 62, 2022

2022
[38]

The Generative Quantum Eigensolver (GQE) and Its Application for Ground State Search,

Kouhei Nakaji et al., “The Generative Quantum Eigensolver (GQE) and Its Application for Ground State Search,” arXiv preprint arXiv:2401.09253, 2024

2024
[39]

Quantum architecture search via deep reinforcement learning,

En-Jui Kuo, Yao-Lung L Fang, and Samuel Yen-Chi Chen, “Quantum architecture search via deep reinforcement learning,” arXiv preprint arXiv:2104.07715, 2021

2021 arXiv
[40]

Quantum architecture search via continual reinforcement learning,

Esther Ye and Samuel Yen-Chi Chen, “Quantum architecture search via continual reinforcement learning,” arXiv preprint arXiv:2112.05779, 2021

2021 arXiv
[41]

Quantum Reinforcement Learning for Quantum Architecture Search,

Samuel Yen-Chi Chen, “Quantum Reinforcement Learning for Quantum Architecture Search,” in International Workshop on Quantum Classical Cooperative, 2023, pp. 17–20

2023
[42]

Quantum Machine Learning Architecture Search via Deep Reinforce- ment Learning,

Xin Dai, Tzu-Chieh Wei, Shinjae Yoo, and Samuel Yen-Chi Chen, “Quantum Machine Learning Architecture Search via Deep Reinforce- ment Learning,” arXiv preprint arXiv:2407.20147 , 2024

2024 arXiv
[43]

Differentiable quantum architecture search for quantum reinforcement learning,

Yize Sun, Yunpu Ma, and V olker Tresp, “Differentiable quantum architecture search for quantum reinforcement learning,” in 2023 IEEE International Conference on Quantum Computing and Engineering . IEEE, 2023, vol. 2, pp. 15–19

2023
[44]

Differentiable Quantum Architecture Search in Asynchronous Quantum Reinforcement Learning,

Samuel Yen-Chi Chen, “Differentiable Quantum Architecture Search in Asynchronous Quantum Reinforcement Learning,” arXiv preprint arXiv:2407.18202, 2024

2024 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [17]

Generalization In Quantum Machine Learning From Few Training Data,

Matthias C Caro et al., “Generalization In Quantum Machine Learning From Few Training Data,” Nature Communications, vol. 13, no. 1, pp. 4919, 2022

work page 2022

[2] [18]

Quantum Machine Learning in Feature Hilbert Spaces,

Maria Schuld and Nathan Killoran, “Quantum Machine Learning in Feature Hilbert Spaces,” Physical Review Letters , vol. 122, no. 4, pp. 040504, 2019

work page 2019

[3] [1]

Li Deng, Deep Learning in Natural Language Processing , Springer, 2018

work page 2018

[4] [2]

Richard Szeliski, Computer Vision: Algorithms and Applications , Springer Nature, 2022

work page 2022

[5] [3]

AI-Powered Drug Discovery Captures Pharma Interest,

Eric Smalley, “AI-Powered Drug Discovery Captures Pharma Interest,” Nature Biotechnology, vol. 35, no. 7, pp. 604–606, 2017

work page 2017

[6] [4]

Michael A Nielsen and Isaac L Chuang, Quantum Computation and Quantum Information, Cambridge University Press, 2010

work page 2010

[7] [5]

Fault-Tolerant Logical Gates in The IBM Quantum Experience,

Robin Harper and Steven T Flammia, “Fault-Tolerant Logical Gates in The IBM Quantum Experience,” Physical review letters , vol. 122, no. 8, pp. 080504, 2019

work page 2019

[8] [6]

CUDA Quantum: The Platform for Integrated Quantum-Classical Computing,

Jin-Sung Kim, Alex McCaskey, Bettina Heim, Manish Modani, Sam Stanwyck, and Timothy Costa, “CUDA Quantum: The Platform for Integrated Quantum-Classical Computing,” in 2023 60th ACM/IEEE Design Automation Conference . IEEE, 2023, pp. 1–4

work page 2023