Pith. sign in

REVIEW 5 major objections 4 minor 53 references

Quantum Classifiers with Trainable Kernel

T0 review · 5 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that training the quantum feature map and classifying from support-vector-only trial states amplifies the decision value and lifts quantum classifiers out of the poor-distinguishability regime of least-squares quantum…

desk verdict A useful but overclaimed QML paper: the support-vector trial state idea is sound and the IRIS numerics are honest, but the theoretical guarantees—especially the multi-class readout bound—are not supported by the analysis. read the letter →

arxiv 2505.04234 v1 pith:2EFTYYZS submitted 2025-05-07 quant-ph

Li Xu , Xiao-yu Zhang , Ming Li , Shu-qian Shen This is my paper · ORCID
classification quant-ph
keywords quantummachinelearningsupportvectortrainablefeaturemappingkernelmethodsamplitudeamplificationmulti-classclassificationdistinguishabilityvariationalcircuits
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

Quantum kernel classifiers lose separation as the training set grows, because the decision value $f(x)=\sum_{i=1}^{M}\alpha_i y_i k(x_i,x)/\sqrt{M}$ shrinks like $O(1/\sqrt{M})$, so the sign is easily buried in measurement noise. This paper proposes two coupled fixes: a trainable quantum feature mapping $U(x,\theta)$ whose parameters are optimized so that same-class samples cluster, and a support-vector-based quantum SVM (SV-QSVM) that builds the trial state only from support vectors, replacing the $1/\sqrt{M}$ normalization with $1/\sqrt{m_s}$. The paper claims this amplifies the decision value and improves both accuracy and distinguishability, and it verifies the gain numerically against the least-squares quantum SVM. It then extends the same amplification idea to a quantum iterative multi-classifier, using amplitude-amplification iteration to read out one-versus-one and one-versus-rest results with high probability. A reader should care because the bottleneck addressed—readout of a tiny inner-product sign—is the step that dominates practical quantum kernel classification.

What carries the argument

The load-bearing objects are the trainable quantum feature mapping $U(x,\theta)=U_l(x,\theta_l)U_{\mathrm{ent}}\cdots U_2(x,\theta_2)U_{\mathrm{ent}}U_1(x,\theta_1)$, a data re-uploading circuit optimized by minimizing the loss $E(\theta)=1-\frac{1}{L}\sum_{j=1}^{L}\frac{1}{M_j}\sum_{i=1}^{M_j}|\langle\psi(x_i^j,\theta)|y_j\rangle|^2$, and the partially evenly weighted trial state $|\upsilon\rangle=\sum_i c_i |i-1\rangle|\psi(x,\theta^*)\rangle/\sqrt{m_s}$ over support vectors, built from the even superposition by amplitude amplification. The training loss pulls same-class feature states toward one label vector per class, which makes the kernel matrix well-conditioned for downstream SVM training. The partial trial state is the mechanism that turns the decision-value shrinkage into an amplification: replacing the $1/\sqrt{M}$ normalization of the least-squares quantum SVM with $1/\sqrt{m_s}$ increases $|f(x)|$ and moves values away from zero, where the sign readout is fragile. For multi-class readout the same amplitude-amplification operator iterates $R\in[1,(\sqrt{5M_j}\pi-1)/2]$ times to concentrate amplitude on the class with the largest overlap, converting a probability that starts near the uniform $1/L$ into one above $0.5$.

What would settle it

Compute the class-wise overlap $a=\sum_i \alpha_i k(x_i,x)/\sqrt{M_j}$ for held-out samples of a trained mapping; if any correctly classified sample has $a<1/\sqrt{5M_j}$ yet the multi-class circuit still reads out correctly after the prescribed $R$ iterations, the Appendix D bound is not the operative mechanism. If every such sample has $a$ below that bound and readout fails, the paper's multi-class readout claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that the poor distinguishability of quantum kernel classifiers on large datasets is not an unavoidable feature of quantum kernels but can be engineered away. The least-squares quantum SVM treats every training sample as a support vector, so its trial state is an evenly weighted superposition over all $M$ samples and its decision value $f(x)=\sum_i\alpha_i y_i k(x_i,x)/\sqrt{M}$ shrinks as $O(1/\sqrt{M})$ by the central limit theorem. SV-QSVM instead learns the support-vector set and classifies with a partially evenly weighted trial state $|\upsilon\rangle=\sum_i c_i |i-1\rangle|\psi(x,\theta^*)\rangle/\sqrt{m_s}$, where $c_i$ marks support vectors; this raises the trial amplitude from $1/\sqrt{M}$ to $1/\sqrt{m_s}$ and therefore amplifies $f(x)$. Equipped with a trainable quantum feature mapping whose parameters $\theta^*$ are optimized to cluster same-class feature states, the paper claims the resulting classifier is strictly more distinguishable than the least-squares quantum SVM and numerically achieves higher accuracy. The paper also claims that iterative amplitude amplification on the stored class overlaps makes multi-class readout reliable with few measurements, provided each class overlap is at least about $1/\sqrt{5M_j}$.

Load-bearing premise

The multi-class readout advantage rests on the assumption that after training, the summed kernel overlap for the true class is at least about $1/\sqrt{5M_j}$, which the paper derives by treating squared kernel overlaps as independent random variables with mean $1/3$; if trained kernels yield smaller overlaps, the prescribed number of amplitude-amplification iterations will not push the readout probability above $0.5$.

Editorial extensions

If this is right

  • A trained TQFM kernel matrix should outperform the untrained kernel matrix for any kernel-based classifier, because the training objective directly separates classes in feature space.
  • SV-QSVM's decision value scales as $1/\sqrt{m_s}$ instead of $1/\sqrt{M}$, so the measurement shots needed to fix the sign of $f(x)$ drop whenever support vectors are a small fraction of the dataset.
  • Total kernel-estimation shots of $O(M^4/\epsilon^2)$ suffice to keep the noisy classifier within $\epsilon$ of the noiseless one, giving a concrete resource bound for the whole SV-QSVM pipeline.
  • Amplitude-amplification iteration over stored class overlaps lets a multi-class label be read out with probability above $0.5$, reducing the readout burden relative to measuring every pairwise overlap separately.
  • The same trained feature mapping can serve as an explicit classifier, an ensemble preparation, and a kernel-matrix source, so the optimization cost is amortized across several classification strategies.

Reading between the lines

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

  • The $O(1/\sqrt{M})$ collapse of the evenly weighted decision value is a generic property of any quantum classifier that superposes all training samples; if the paper's diagnosis is right, the support-vector subsetting trick could be applied to other kernel-based quantum algorithms, not just SVMs.
  • The Appendix D bound assumes squared kernel overlaps have mean $1/3$ and behave like independent random variables; this is testable directly by histogramming $k(x_i,x)^2$ on a trained mapping, and the multi-class advantage would be on firmer ground if the same readout worked when that bound fails.
  • One could replace the amplitude-amplification iteration with quantum amplitude estimation to estimate each class overlap to a chosen precision, trading the threshold assumption for standard phase-estimation overhead.
  • The numerical comparison is on a small, well-clustered benchmark; the claimed accuracy gap should be probed on larger and noisier kernels before treating the distinguishability gain as universal.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 4 minor

Summary. The paper proposes a trainable quantum feature mapping (TQFM) built from data re-uploading circuits, and uses it in three ways: as an explicit classifier, as an ensemble model, and as a source of quantum kernels for support vector machines. The main algorithmic contribution is the SV-QSVM, which constructs classification trial states from support vectors only, rather than from all training samples, in order to improve the distinguishability of the decision value. The paper also introduces a quantum iterative multi-classifier based on amplitude amplification to read out one-versus-one or one-versus-rest results, with a proposed iteration bound in Appendix D. Numerical simulations on the IRIS dataset are reported for each component, showing improved clustering and classification accuracy compared to a standard quantum feature map and to the LS-QSVM baseline.

Significance. If the theoretical claims were rigorously established, the SV-QSVM construction would be a useful and intuitive contribution to quantum kernel methods: restricting the trial state to support vectors is a natural way to amplify the decision value and could reduce measurement overhead. The numerical demonstrations on IRIS support the qualitative story, and the paper is honest about the simplicity of the experiments. The work also points to an interesting direction in using amplitude amplification for multi-class readout. However, several of the paper's central theoretical justifications—Theorem 1, Lemma 2, and the Appendix D iteration bound—contain gaps that prevent the results from being accepted as rigorous. The empirical results are promising but do not by themselves supply the missing proofs.

major comments (5)
  1. [Section III C, Theorem 1] The proof of Theorem 1 assumes that the kernel values k(x_i,x) are independent and identically distributed and then invokes the Central Limit Theorem to conclude f(x) ~ N(0, 1/M). This is not justified: for a fixed trained feature map, the k(x_i,x) are not independent, and more importantly they are nonnegative (or at least not zero-mean in general), so the CLT limit has a positive mean rather than zero mean. Consequently the claimed scaling f(x) = O(1/sqrt(M)) is not established. This theorem is load-bearing because it motivates the entire SV-QSVM construction as a fix for poor distinguishability. The statement should be rephrased as a conditional statement under explicit assumptions on the kernel-value distribution, or replaced by a rigorous bound.
  2. [Appendix D and Section V B] The lower bound a ≈ 1/sqrt(5 M_j) on the target amplitude is derived by modeling g(x) = sum_i alpha_i k(x_i,x)/sqrt(M_j) as a normal N(0, 1/(3 M_j)) and then writing ∫_0^a g(x) dx = 1/4. This is not a valid probability calculation: g(x) is a random variable, not a probability density, and the zero-mean normal model is inconsistent with k(x_i,x) ∈ [0,1] because the mean of a sum of nonnegative overlaps is nonnegative. If the true mean is positive, a tends to a constant rather than to 0 as M_j grows, so the iteration bound R ∈ [1, (sqrt(5 M_j) π − 1)/2] is not a worst-case bound. If the mean is very small, a can be smaller than 1/sqrt(5 M_j) and the prescribed number of Grover iterations is insufficient. Thus the claim that the quantum iterative multi-classifier reads out results with high probability is unsupported by the stated analysis.
  3. [Section IV D, Lemma 2] The proof of Lemma 2 is incomplete. The bound ∥δ∥_2 ≤ O(M/sqrt(R)) introduces a quantity R that is never defined, and the step from the perturbation bound in Lemma 1 to this expression is not shown. The argument that λ_min(K + (1/γ)I) ≥ 1/γ is a 'constant lower bound' is true for the noiseless matrix, but Lemma 1 applies to the noisy matrix K', and the effect of noise on the eigenvalue should be stated explicitly. The final shot count O(M^4/ε^2) therefore does not follow from the given reasoning. This lemma is important for the claimed error analysis of SV-QSVM, so the proof needs to be made rigorous or the claim weakened.
  4. [Section IV D, misclassification-rate analysis] The derivation of the misclassification rate is internally inconsistent. The text first defines p(x) as the probability density of h(x) and states ∫_{-1}^{1} p(x) dx = 1, but then writes the maximum misclassification rate as ∫_0^ε p(x) dx, and later uses integrals such as 1 − ∫_{-ε}^{ε} p(x) dx and ∫_{-ε}^{-ε} p(x) dx. Since x is the trial datum and h(x) is a function of x, the integral limits in x-space do not correspond to decision errors in h-space. The probabilistic model needs to be clarified, for example by defining the distribution of h(x) directly and expressing the error probability as an integral over the decision threshold.
  5. [Section VI A, closing paragraph] The sentence 'using the trained kernel matrix for machine learning algorithms, such as SVM, will definitely outperform the untrained one in terms of performance' is an overclaim. The simulations demonstrate improvement on the IRIS dataset for one specific circuit layout, but they do not establish a general guarantee. This statement should be rephrased as an empirical observation about the tested cases.
minor comments (4)
  1. [Section IV D] The phrase 'partially evenly weighted trail state' appears to be a typo for 'trial state'; the same term is used correctly elsewhere in the paper.
  2. [Section IV D, Lemma 2] The symbol R is used in the proof without definition; if it denotes the number of measurement shots, this should be stated explicitly and consistently with the rest of the paper.
  3. [Table II] In the row for Class 2 with r=1, the average probability after iteration is printed as '0747', which appears to be a typo for '0.747'.
  4. [Section V B and Appendix D] The preparation of the oracles U_L and U_x and the state U_x|j-1⟩|0...0⟩ assumes the ability to coherently superpose training samples and the new datum; the resource cost of these oracles is not discussed, and this assumption should be stated explicitly when claiming a reduced readout burden.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: central claims are derived from explicit constructions, not from fitted inputs or load-bearing self-citations.

full rationale

The paper's derivation chain is self-contained. The TQFM parameters are trained by minimizing Eq. (2), a loss function that directly uses label overlaps, and the trained kernel is then evaluated on new data; no fitted quantity is renamed as a prediction. The SV-QSVM decision function in Section IV C follows from explicit state definitions |upsilon> and |mu>, with the support-vector restriction and the 1/sqrt(m_s) normalization part of the construction. The claimed distinguishability advantage is a direct consequence of the normalization 1/sqrt(m_s) versus 1/sqrt(M) and the concentration argument in Section IV D, not a reduction of the claim to its inputs. The multi-classifier iteration bound in Appendix D is statistically questionable because kernel overlaps are nonnegative while the derivation assumes a zero-mean normal model; however this is a correctness/robustness weakness, not a circular step, since the bound is an assumed probabilistic model rather than a quantity fitted to the very prediction it supports. The only self-citation, Ref. [42], is used for a robustness inequality and as a variational implementation alternative in the numerics; it does not carry the central claims and is not load-bearing. Under the rubric, self-citation alone does not constitute circularity. Overall, no step reduces by construction to its own inputs.

Assumptions & free parameters 6 free parameters · 7 assumptions · 0 invented entities

The central claims rest on standard SVM and kernel mathematics plus several statistical assumptions about kernel-value distributions, and on an ad hoc 1/3 second-moment assumption in Appendix D. The method itself adds trainable parameters and hyperparameters that are fit to the IRIS data during simulations.

free parameters (6)
  • TQFM parameters theta = theta* from COBYLA optimization on IRIS training data
    Appear in U(x,theta), optimized by minimizing Eq. (2); values depend on layer count r and rotation choice and are not reported.
  • SVM dual coefficients alpha = Optimized in simulations, exact values not reported
    Solved via the variational objective Eq. (6); used to construct the partial weighted trial state.
  • Regularization constant gamma
    Appears in Eq. (4) and Eq. (6) as SVM regularization; no value is given.
  • Penalty multiplier C
    Added in Eq. (6) as a penalty for the equality constraint; no value is given.
  • Support-vector threshold
    Data-dependent cutoff for identifying support vectors from measurement frequencies in Section IV C; threshold is not specified.
  • Layer count r and rotation type = r = 0, 1, 2; R_y or R_z-R_y-R_z
    Chosen by hand in the simulations; results vary with r.
assumptions (7)
  • standard math Soft-margin SVM dual formulation and kernel matrix positive semidefiniteness.
    Used in Section IV A to set up Eq. (5) and to justify Lemma 2's eigenvalue lower bound.
  • domain assumption Kernel values k(x_i,x) for a random trial x are independent and identically distributed across training samples.
    Theorem 1 assumes this to apply the Central Limit Theorem; not guaranteed for structured data.
  • domain assumption The Central Limit Theorem can be applied to sums of kernel values and to squared kernel overlaps.
    Used in Theorem 1 and Appendix D; requires finite variance and independence, which are asserted rather than proven.
  • standard math Quantum kernel entries can be estimated to precision epsilon' with O(M^2 / epsilon'^2) shots via the inversion or SWAP test.
    Taken from Refs [16,30]; used in Lemma 2 to derive total shot complexity.
  • domain assumption The TQFM circuit U(x,theta) is efficiently implementable and trainable without suffering from barren plateaus.
    The authors invoke local loss functions to avoid barren plateaus but provide no formal guarantee for their specific layout.
  • ad hoc to paper Squared kernel overlaps have mean 1/3, so the multi-class target amplitude is at least about 1/sqrt(5 M_j).
    Stated in Appendix D with no derivation; this assumption drives the Grover iteration count.
  • domain assumption Support-vector locations can be recovered from measurement frequencies of U_A(xi*) using a threshold.
    Section IV C; the threshold is unspecified and errors below threshold are assumed to be noise.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum Classifiers with Trainable Kernel." pith.science (2026). https://pith.science/paper/2EFTYYZS

@misc{pith2026250504234,
  author       = {Pith},
  title        = {Pith review of: Quantum Classifiers with Trainable Kernel},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2EFTYYZS}},
  note         = {Machine review of arXiv:2505.04234}
}
read the original abstract

Kernel function plays a crucial role in machine learning algorithms such as classifiers. In this paper, we aim to improve the classification performance and reduce the reading out burden of quantum classifiers. We devise a universally trainable quantum feature mapping layout to broaden the scope of feature states and avoid the inefficiently straight preparation of quantum superposition states. We also propose an improved quantum support vector machine that employs partially evenly weighted trial states. In addition, we analyze its error sources and superiority. As a promotion, we propose a quantum iterative multi-classifier framework for one-versus-one and one-versus-rest approaches. Finally, we conduct corresponding numerical demonstrations in the \textit{qiskit} package. The simulation result of trainable quantum feature mapping shows considerable clustering performance, and the subsequent classification performance is superior to the existing quantum classifiers in terms of accuracy and distinguishability.

Figures

Figures reproduced from arXiv: 2505.04234 by the authors.

Figure 1
Figure 1. FIG. 1. (a) Circuits representation of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Hadamard test for calculating [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Probability density functions in several different cases. The black dashed line represents the error between a noisy classifier and a [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Numerical verification of trainable quantum feature mapping. (a)A basic TQFM framework used in the simulations. (b) The iteration [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Sample density distribution where the function value is [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Circuit diagram of quantum iterative multi-classifier with 1 [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 45 canonical work pages

  1. [1]

    Shor, Polynomial-Time Algorithms for Prime Factorization and Discrete Logarithms on a Quantum Computer, SIAM Jour- nal on Computing, 26, 1484(1997)

    P. Shor, Polynomial-Time Algorithms for Prime Factorization and Discrete Logarithms on a Quantum Computer, SIAM Jour- nal on Computing, 26, 1484(1997)

    work page 1997

  2. [2]

    L. K. Grover, Quantum Mechanics Helps in Searching for a Needle in a Haystack, Phys. Rev. Lett. 79, 325 (1997)

    work page 1997

  3. [3]

    A. W. Harrow , A. Hassidim, and S. Lloyd, Quantum algorithm for linear systems of equations, Phys. Rev. Lett. 103, 150502 (2009)

    work page 2009

  4. [4]

    Biamonte, P

    J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. wiebe, and S. Lloyd, Quantum machine learning, Nature 549, 195-202 (2017)

    work page 2017

  5. [5]

    Dunjko and H

    V . Dunjko and H. J. Briegel, Machine learning & artificial in- telligence in the quantum domain: a review of recent progress, Rep. Prog. Phys. 81,074001(2018)

    work page 2018

  6. [6]

    Lloyd, M

    S. Lloyd, M. Mohseni, and P. Rebentrost, Quantum principal component analysis, Nat. Phys. 10, 631-633 (2014)

    work page 2014

  7. [7]

    W. Ren, W. Li, S. Xu, K. Wang, W. Jiang, F. Jin, X. Zhu, J. Chen, Z. Song, P. Zhang, et al., Experimental quantum adver- sarial learning with programmable superconducting qubits, Nat. Comput. Sci. 2, 711-717 (2022)

    work page 2022

  8. [8]

    Cortes and V

    C. Cortes and V . Vapnik, Support-vector networks, Mach. Learn. 20, 273-297 (1995)

    work page 1995

Show all 53 references
  1. [9]

    Vapnik,Statistical learning theory, 1998 (V ol

    V . Vapnik,Statistical learning theory, 1998 (V ol. 3). New York, NY:Wiley, 1998: Chapter 10-11, pp.401-492

    1998

  2. [10]

    Rebentrost, M

    P. Rebentrost, M. Mohseni, and S. Lloyd, Quantum support vec- tor machine for big data classification, Phys. Rev. Lett. 113, 130503 (2014)

    2014

  3. [11]

    Li, X.-M

    Z.-K. Li, X.-M. Liu, N.-Y . Xu, and J.-F. Du, Experimental real- ization of a quantum support vector machine, Phys. Rev. Lett. 114, 110504 (2015)

    2015

  4. [12]

    Giovannetti, S

    V . Giovannetti, S. Lloyd, and L. Maccone, Quantum random access memory, Phys. Rev. Lett. 100, 160501 (2008)

    2008

  5. [13]

    Giovannetti, S

    V . Giovannetti, S. Lloyd, and L. Maccone, Architectures for a quantum random access memory, Phys. Rev. A 78, 052310 (2008)

    2008

  6. [14]

    Y . R. Sanders, G. H. Low, A. Scherer, and D. W. Berry, Black- box quantum state preparation without arithmetic, Phys. Rev. Lett. 122, 020502 (2019)

    2019

  7. [15]

    Schuld and N

    M. Schuld and N. Killoran, Quantum machine learning in fea- ture Hilbert spaces, Phys. Rev. Lett. 122, 040504 (2019)

    2019

  8. [16]

    Havl´ıˇ cek, A

    V . Havl´ıˇ cek, A. D. C ´orcoles, K. Temme, A. W. Harrow, A. Kandala, J. M. Chow, and J. M. Gambetta, Supervised learning with quantum-enhanced feature spaces, Nature 567, 209-212 (2019)

    2019

  9. [17]

    Y . Liu, S. Arunachalam, and K. Temme, A rigorous and robust quantum speed-up in supervised machine learning, Nat. Phys. 17, 1013-1017 (2021)

    2021

  10. [18]

    Blank, D

    C. Blank, D. K. Park, J-K. K. Rhee, and F. Petruccione, Quan- tum classifier with tailored quantum kernel, npj Quantum Inf. 6, 41 (2020)

    2020

  11. [19]

    J ¨ager and R

    J. J ¨ager and R. V . Krems, Universal expressiveness of varia- tional quantum classifiers and quantum kernels for support vec- tor machines, Nat. Commun. 14, 576 (2023)

    2023

  12. [20]

    T. Goto, Q. H. Tran, and K. Nakajima, Universal approxima- tion property of quantum machine learning models in quantum- enhanced feature spaces, Phys. Rev. Lett. 127, 090506 (2021)

    2021

  13. [21]

    D. K. Park, C. Blank, and F. Petruccione, Te theory of the quan- tum kernel-based binary classifer, Phys. Lett. A 384, 126422

  14. [22]

    P ´erez-Salinas, A

    A. P ´erez-Salinas, A. Cervera-Lierta, E. Gil-Fuster, and J. I. Latorre, Data re-uploading for a universal quantum classifier, Quantum 4, 226(2020)

    2020

  15. [23]

    Jerbi, L.J

    S. Jerbi, L.J. Fiderer, H. Poulsen Nautrup, J. M. K ¨ubler, H. J. Briegel, and V . Dunjko, Quantum machine learning beyond ker- nel methods. Nat. Commun. 14, 517 (2023)

    2023

  16. [24]

    Cincio, Y

    L. Cincio, Y . Subas, A. T. Sornborger, and P. JColes, Learn- ing the quantum algorithm for state overlap, New J. Phys. 20,113022 (2018)

    2018

  17. [25]

    Killoran, T

    N. Killoran, T. R. Bromley, J. M. Arrazola, M. Schuld, N. Que- sada, and S. Lloyd, Continuous-variable quantum neural net- works, Phys. Rev. Research 1, 033063 (2019)

    2019

  18. [26]

    Du, M.-H

    Y . Du, M.-H. Hsieh, T. Liu, S. You, and D. Tao, Learnabil- ity of Quantum Neural Networks, PRX QUANTUM 2, 040337 (2021). 13

    2021

  19. [27]

    Y . Du, Y . Yang, D. Tao, and M.-H. Hsieh, Problem-Dependent Power of Quantum Neural Networks on Multiclass Classifica- tion, Phys. Rev. Lett. 131, 140601 (2023)

    2023

  20. [28]

    W. Li, Z. Lu, and D.-L. Deng, Quantum neural network classi- fiers: A tutorial, SciPost Phys. Lecture Notes, Aug. 2022, Art. no. 61

    2022

  21. [29]

    X. Pan, Z. Lu, W. Wang, Z. Hua, Y . Xu, W. Li, W. Cai, X. Li, H. Wang, Y .-P. Song,et al., Deep quantum neural networks on a superconducting processor. Nat. Commun. 14, 4006 (2023)

    2023

  22. [30]

    Lloyd, M

    S. Lloyd, M. Schuld, A. Ijaz, J. Izaac, and N. Killoran, Quantum embeddings for machine learning, arXiv:2001.03622

    2001 arXiv

  23. [31]

    N. A. Nghiem, S. Y .-C. Chen, and T.-C. Wei, Unified frame- work for quantum classification, Phys. Rev. Research 3, 033056(2021)

    2021

  24. [32]

    Peruzzo, J

    A. Peruzzo, J. McClean, P. Shadbolt, Man-Hong Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O’Brien, A varia- tional eigenvalue solver on a photonic quantum processor, Nat. Commun. 5, 4213 (2014)

    2014

  25. [33]

    Kandala, A

    A. Kandala, A. Mezzacapo, K. Temme, M. Takita, M. Brink, J. M. Chow, and J. M. Gambetta, Hardware-efficient variational quantum eigensolver for small molecules and quantum mag- nets, Nature 549, 242-246 (2017)

    2017

  26. [34]

    Nam, J.-S

    Y . Nam, J.-S. Chen, N. C. Pisenti, K. Wright, C. Delaney, D. Maslov, K. R. Brown, S. Allen, J. M. Amini, J. Apisdorf, et al., Ground-state energy estimation of the water molecule on a trapped-ion quantum computer, npj Quantum Inf. 6, 33 (2020)

    2020

  27. [35]

    D. Wang, O. Higgott, S. Brierley, Accelerated variational quan- tum eigensolver, Phys. Rev. Lett. 122, 140504 (2019)

    2019

  28. [36]

    R. M. Parrish, E. G. Hohenstein, P. L. McMahon, and T. J. Mart´ınez, Quantum computation of electronic transitionsUsing a variational quantum eigensolver, Phys. Rev. Lett. 122, 230401 (2019)

    2019

  29. [37]

    W. M. Kirby and P. J. Love, Variational quantum eigensolvers for sparse hamiltonians, Phys. Rev. Lett. 127, 110503 (2021)

    2021

  30. [38]

    X. Xu, J. Sun, S. Endo, Y . Li, S. C. Benjamin, and X. Yuan, Variational algorithms for linear algebra, Sci. Bull. 66, 2181- 2188 (2021)

    2021

  31. [39]

    X. Wang, Z. Song, and Y . Wang, Variational Quantum Singular Value Decomposition, Quantum 5, 483 (2021)

    2021

  32. [40]

    Jones, M

    J. Jones, M. Mosca, and R. Hansen, Implementation of a quan- tum search algorithm on a quantum computer, Nature 393, 344- 346 (1998)

    1998

  33. [41]

    G. L. Long, Grover algorithm with zero theoretical failure rate, Phys. Rev. A, 64:022307(2001)

    2001

  34. [42]

    Xu, X.-Y

    L. Xu, X.-Y . Zhang, J.-M. Liang, J. Wang, M. Li, L. Jian, and S.-Q. Shen, Variational quantum support vector machine based on Hadamard test, Commun. Theor. Phys. 74 055106(2022)

    2022

  35. [43]

    J. W. Daniel, Stability of the solution of definite quadratic pro- grams, Mathematical Programming 5, 41 (1973)

    1973

  36. [44]

    Koczor and S

    B. Koczor and S. C. Benjamin, Quantum analytic descent, Phys. Rev. Research 4, 023017 (2022)

    2022

  37. [45]

    Schuld, V

    M. Schuld, V . Bergholm, C. Gogolin, J. Izaac, and N. Killoran, Evaluating analytic gradients on quantum hardware, Phys. Rev. A 99, 032331 (2019)

    2019

  38. [46]

    Sheng and L

    Y .-B. Sheng and L. Zhou, Distributed secure quantum machine learning. Sci. Bull. 62, 1025–1029 (2017)

    2017

  39. [47]

    Li and D.-L

    W. Li and D.-L. Deng, Recent advances for quantum classifiers, Sci. China Phys. Mech. Astron. 65, 220301 (2022)

    2022

  40. [48]

    McClean, S

    J.R. McClean, S. Boixo, V . N. Smelyanskiy, R. Babbush,and H. Neven, Barren plateaus in quantum neural network training landscapes. Nat. Commun. 9, 4812 (2018)

    2018

  41. [49]

    Cerezo, A

    M. Cerezo, A. Sone, T. V olkoff, L. Cincio, and P. J. Coles, Cost function dependent barren plateaus in shallow parametrized quantum circuits, Nat. Commun. 12, 1791 (2021)

    2021

  42. [50]

    Cerezo, M

    M. Cerezo, M. Larocca, D. Garc ´ıa-Mart´ın, N. L. Diaz, P. Braccia, E. Fontana, M. S. Rudolph, P. Bermejo, A. Ijaz, S. Thanasilp, et al., Does provable absence of barren plateaus im- ply classical simulability? Or, why we need to rethink varia- tional quantum computing. arXiv:...

    2023 arXiv

  43. [51]

    M. W. Johnson, M. H. S. Amin, S. Gildert, T. Lanting, F. Hamze, N. Dickson, R. Harris, A. J. Berkley, J. Johansson, P. Bunyk, et al., Quantum annealing with manufactured spins. Na- ture 473, 194-198(2011)

    2011

  44. [52]

    Gibney, D-Wave upgrade: How scientists are using the world’s most controversial quantum computer, Nature 541, 447-448(2017)

    E. Gibney, D-Wave upgrade: How scientists are using the world’s most controversial quantum computer, Nature 541, 447-448(2017)

    2017

  45. [53]

    Willsch, M

    D. Willsch, M. Willsch, H. De Raedt, and K. Michielsen, Sup- port vector machines on the d-wave quantum annealer, Comput. Phys. Commun. 248, 107006(2020)

    2020

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.