Pith. sign in

REVIEW 4 major objections 5 minor 43 references

QSVM-QNN: Quantum Support Vector Machine Based Quantum Neural Network Learning Algorithm for Brain-Computer Interfacing Systems

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A hybrid QSVM-QNN quantum classifier is tested on two EEG datasets, but its reported accuracies are not superior to classical baselines in the paper's own tables.

desk verdict The abstract says QSVM-QNN beats classical and quantum baselines; its own Table I shows it doesn't, and the statistics are invalid. read the letter →

arxiv 2505.14192 v1 pith:MHNNVLUT submitted 2025-05-20 quant-ph cs.LGeess.SP

Bikash K. Behera , Saif Al-Kuwari , Ahmed Farouk This is my paper · ORCID
classification quant-phcs.LGeess.SP
keywords quantumqsvm-qnnclassificationlearningmodelsqsvmsystemsbrain-computer
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

Brain-computer interfaces read electrical signals from the scalp and try to work out what the person is doing or thinking. This paper tries to classify such EEG signals with a hybrid quantum model, called QSVM-QNN. The QSVM part is not a real support vector machine: it uses k-means to split the training data into two clusters, then encodes each new test point and the two cluster centers into a small quantum circuit and reads out how much the test point overlaps with each center. The QNN part adds a layer of adjustable quantum rotations, trained with the COBYLA optimizer, to improve the labels. Everything runs on a simulator with 1024 shots per circuit.

The authors report accuracies of 0.990 and 0.950 on two Kaggle EEG datasets and claim the model outperforms classical and standalone quantum models. The paper's own Table I does not support that claim: on the first dataset, logistic regression, CNN and SVM all score higher than 0.990, and on the second dataset almost every classical model plus the standalone QSVM beats 0.950. The statistical tests are also not meaningful: each algorithm has only two numbers, one per dataset, with no repeated runs, no variance, and no error bars, and the text says the QSVM comparison is non-significant while the table reports p-values around 10^-21. No code, data, or circuit specifications are released.

Extended reading notes

Core claim

Abstract: 'The proposed model is evaluated on two benchmark EEG datasets, achieving high accuracies of 0.990 and 0.950, outperforming both classical and standalone quantum models.' If correct, QSVM-QNN would be a noise-robust hybrid quantum classifier that beats classical ML on EEG BCI tasks.

Load-bearing premise

Section III.C (Eq. 22-28) and Algorithms 2-3 assume the k-means centroids computed in PCA-reduced 3-qubit space represent the two EEG classes well enough that the overlap <0|U(c)†U(x)|0>, read as sqrt of the all-zero measurement probability, is a valid decision signal. The paper provides no evidence that these centroids capture class structure, and the method is not a support vector machine in the sense of Eq. (10)-(11).

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.

Referee Report

4 major / 5 minor

Summary. The paper proposes a hybrid quantum-classical classifier, QSVM-QNN, for binary EEG-based brain-computer interface tasks, combining a centroid-overlap "QSVM" circuit (the U-U-dagger method) with a variational quantum circuit inspired by QNN. The authors evaluate the model on two Kaggle EEG datasets against classical baselines (KNN, RF, SVM, CNN, LR, DT, NB) and quantum baselines (QNN, QSVM), and report noise robustness under six error models. The central claim is that QSVM-QNN achieves accuracies of 0.990 and 0.950 and outperforms both classical and standalone quantum models; the paper also reports ANOVA and t-test results to support the comparisons.

Significance. If the central claim were correct, a quantum classifier that is both more accurate than classical models and robust to realistic noise would be a notable contribution for BCI applications. The manuscript has some positive features: it presents explicit circuit diagrams and pseudocode for all three quantum algorithms, evaluates two datasets, and considers six noise models. However, the headline superiority claim is contradicted by the paper's own Table I, the statistical tests are built on an invalid sample size, and the algorithm labeled QSVM is not a support vector machine in the sense of the paper's own equations. These issues are load-bearing: the abstract, introduction, and conclusion all rest on the refuted outperformance claim. The significance of the work, as currently presented, is therefore low, and the experimental evidence does not support the stated conclusions.

major comments (4)
  1. [Abstract, Section I.A, Section IV.F, Table I] The abstract's claim that QSVM-QNN achieves 0.990 and 0.950 accuracy while 'outperforming both classical and standalone quantum models' is directly contradicted by Table I. On EEGMM, LR achieves 0.997, CNN 0.996, and SVM 0.995, all above QSVM-QNN's 0.990. On EEG, SVM, RF, and LR achieve 0.999, KNN and CNN achieve 0.998, and QSVM achieves 0.998, all above QSVM-QNN's 0.950. Section V itself states that 'QSVM performed best in the EEG dataset.' The experimental results therefore do not support the paper's central contribution as stated in the abstract and introduction.
  2. [Section IV.G, Tables II-VI] The statistical significance tests are invalid as conducted. Each model has only two accuracy values in Table I (one per dataset), yet the paper reports t-tests with p-values as small as 2.18e-44 and an ANOVA with p=1.47e-195. These values cannot arise from a valid two-replicate analysis and are not meaningful evidence for performance differences. In addition, the text in Section IV.G states that the QSVM comparison yields 'non-significant p-values across all metrics,' but Table III reports p=1.19e-21 for accuracy, which is highly significant by any conventional threshold. The statistical analysis must be redone with proper sample sizes or, at minimum, the reported p-values and their interpretation must be corrected.
  3. [Section III.C, Eq. (10)-(11), Algorithms 2 and 3] The algorithm called QSVM is not a support vector machine as defined by Eqs. (10)-(11). The method described in Section III.C and Algorithm 2 computes inner products between a test point and two k-means centroids using the U U-dagger circuit, and assigns labels based on the square-root of the all-zero measurement probability. There is no hinge-loss minimization, no learned alpha_i coefficients, and no identification of support vectors. Consequently, the claim that QSVM-QNN integrates 'the decision boundary capabilities of QSVM' is unsupported by the actual method. The paper provides no evidence that k-means centroids in the PCA-reduced 3-qubit feature space capture the class structure needed for reliable classification.
  4. [Section IV.H, Figs. 6-7, Abstract] The noise-robustness conclusion is not established by the presented results. The text in Section IV.H describes the accuracy curves qualitatively, with no error bars, no repeated runs, and no statistical test comparing noisy and noise-free performance. Moreover, Fig. 6(a) indicates that bit-flip and bit-phase-flip noise improve QNN accuracy on the EEGMM dataset as the noise parameter increases, a counterintuitive result that is not explained. Without additional evidence, the abstract's claim that QSVM-QNN 'maintains stable performance under noisy conditions' and is 'noise-resilient' is not supported.
minor comments (5)
  1. [Section IV.F] The text states 'QSVM-QNN has the highest accuracy of 0.997 on the EEGMM dataset,' but Table I reports 0.990 for QSVM-QNN on EEGMM; this inconsistency should be corrected.
  2. [Section II.A, Eq. (22)] The notation 'n⊗ <0|' is malformed; it should presumably read '<0|^{⊗n}' to denote the n-fold tensor product of the bra vector.
  3. [Section IV.A and Section IV.I] Section IV.A says 'one layer of parametrized circuit is used, while 5 epochs of QNN are used,' but Section IV.I says 'only 5 layers of QNN have been taken'; the manuscript should clarify whether the depth is one layer or five layers and distinguish epochs from layers.
  4. [References] Reference [33] contains a typo in the author name ('Macccone' instead of 'Maccone'), and the reference list should be checked for formatting consistency.
  5. [Section I.A] The motivation paragraph says QSVM-QNN 'achieved 0.990 accuracy on the EEGMM dataset and 0.950 accuracy on the EEG dataset, outperforming QNN while being comparable to QSVM,' which is inconsistent with the abstract's stronger claim of outperforming both classical and standalone quantum models; the internal claims should be aligned.

Circularity Check

1 steps flagged · score 2.0 of 10

No definitional circularity: the reported accuracies are empirical train/test simulation outcomes, not quantities defined in terms of the answer. The main circularity-adjacent issue is that the 'QSVM' component is a k-means centroid-overlap classifier renamed as a support-vector machine and imported from the authors' own UU† references.

  1. renaming known result [Section III.C (Eqs. 20-28, Algorithm 2) versus Section II.C (Eqs. 10-11)]
    "Here, we explain the QSVM algorithm, particularly following the U U† method as described in [34]. The primary objective of this algorithm is to compute the inner product between the centroid of a dataset cluster and a given test data point in the quantum feature space. ... Apply KMeans clustering on the dataset features ... Assign predicted label based on probabilities."

    Algorithm 2 assigns labels by comparing overlap probabilities with two k-means centroids; it never computes support vectors, alpha_i coefficients, or the hinge loss of Eqs. (10)-(11). The component called 'QSVM' is therefore a quantum nearest-centroid classifier, not a support-vector machine. The paper's central claim that QSVM-QNN combines 'the decision boundary capabilities of QSVM' with QNN inherits that capability only by renaming the centroid classifier, not from the SVM derivation the paper itself presents.

full rationale

The headline accuracies (0.990 and 0.950) are empirical outcomes of running the circuits and evaluating on a held-out test split (Section IV.F, Table I), so they are not defined in terms of the answer and do not reduce to a fit. The main circularity-adjacent pattern is the renaming of a k-means centroid-overlap classifier as 'QSVM': the paper's own SVM formulation in Eqs. (10)-(11) is not implemented in Algorithm 2, and the UU† method is taken from the authors' prior work ([34], with [35] supplying the probability identity), making the 'QSVM' label a self-referential relabeling rather than a derived SVM. This does not, however, force the numerical predictions: the simulations, noise scans, and comparisons in Table I are independent empirical content. The invalid t-test/ANOVA claims and the Abstract's 'outperforming' statement are correctness problems, not circularity. Overall circularity is low.

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

The central performance claims rest on hand-chosen hyperparameters (3 qubits, 1 VQC layer, 5 QNN epochs, 0.2 test fraction, k-means centroids, unstated PCA dimension) and on the validity of the overlap-probability decision rule. No new physical entities are introduced; the ledger is dominated by modeling assumptions about the EEG features and the classifier.

free parameters (7)
  • VQC rotation angles theta = Optimized per dataset via COBYLA; values not reported
    Central trainable parameters in Eq. (29)-(30); reported accuracies depend on them.
  • Number of qubits = 3
    Section IV.A: 'for quantum algorithms, 3 qubits are used'; chosen by hand, not justified, and it limits the PCA-reduced feature dimension.
  • Number of VQC layers = 1
    Section IV.A: one layer of parametrized circuit is used; authors claim more layers do not improve accuracy but show no supporting data.
  • Number of QNN epochs = 5
    Section IV.A: '5 epochs of QNN are used'; fixed without a reported sensitivity analysis.
  • K-means centroids c1, c2 = Computed per dataset; values not reported
    Algorithms 2 and 3: QSVM and QSVM-QNN classification is fully determined by these centroids and the overlap probabilities.
  • PCA feature count = Not stated (chosen to fit 3-qubit amplitude encoding)
    Section IV.C: 'the number of features is chosen to encode the data points properly'; exact dimension omitted, affecting all results.
  • Noise model parameters = Varied from 0 to 1
    Section IV.A and IV.H: robustness curves are generated by sweeping noise parameters, but accuracy at each setting is not tabulated and no repeated runs are reported.
assumptions (6)
  • standard math Amplitude encoding maps classical features into normalized quantum states; all gates are unitary and measurement follows the Born rule.
    Used throughout Section III; standard quantum mechanics.
  • domain assumption The inner product <0|U(c)†U(x)|0> equals the coefficient alpha_1,0, and alpha_1,0 is real and equal to sqrt(P_000...000).
    Eq. (22)-(28); requires the overlap to be real and non-negative, which is not guaranteed for arbitrary U and limits the classifier.
  • domain assumption K-means centroids computed in the PCA-reduced feature space adequately represent the two classes for the quantum overlap classifier.
    Algorithms 2 and 3; if the centroids are poor, the inner-product decision rule fails.
  • domain assumption PCA-reduced EEG features encoded into 3 qubits retain sufficient class-discriminative information.
    Section IV.A and IV.C; no ablation, reconstruction error, or feature-selection validation is provided.
  • domain assumption The two Kaggle datasets are accurate and representative BCI benchmark data.
    Section IV.B; labels and recordings are taken as ground truth without independent validation.
  • ad hoc to paper Statistical tests treat each dataset's metric as an independent replicate of model performance.
    Section IV.G; n=2 per algorithm, no repeated runs, violating independence and producing absurd p-values.

how reviews work

0 comments
Cite this review

Pith. "Pith review of QSVM-QNN: Quantum Support Vector Machine Based Quantum Neural Network Learning Algorithm for Brain-Computer Interfacing Systems." pith.science (2026). https://pith.science/paper/MHNNVLUT

@misc{pith2026250514192,
  author       = {Pith},
  title        = {Pith review of: QSVM-QNN: Quantum Support Vector Machine Based Quantum Neural Network Learning Algorithm for Brain-Computer Interfacing Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MHNNVLUT}},
  note         = {Machine review of arXiv:2505.14192}
}
read the original abstract

A brain-computer interface (BCI) system enables direct communication between the brain and external devices, offering significant potential for assistive technologies and advanced human-computer interaction. Despite progress, BCI systems face persistent challenges, including signal variability, classification inefficiency, and difficulty adapting to individual users in real time. In this study, we propose a novel hybrid quantum learning model, termed QSVM-QNN, which integrates a Quantum Support Vector Machine (QSVM) with a Quantum Neural Network (QNN), to improve classification accuracy and robustness in EEG-based BCI tasks. Unlike existing models, QSVM-QNN combines the decision boundary capabilities of QSVM with the expressive learning power of QNN, leading to superior generalization performance. The proposed model is evaluated on two benchmark EEG datasets, achieving high accuracies of 0.990 and 0.950, outperforming both classical and standalone quantum models. To demonstrate real-world viability, we further validated the robustness of QNN, QSVM, and QSVM-QNN against six realistic quantum noise models, including bit flip and phase damping. These experiments reveal that QSVM-QNN maintains stable performance under noisy conditions, establishing its applicability for deployment in practical, noisy quantum environments. Beyond BCI, the proposed hybrid quantum architecture is generalizable to other biomedical and time-series classification tasks, offering a scalable and noise-resilient solution for next-generation neurotechnological systems.

Figures

Figures reproduced from arXiv: 2505.14192 by the authors.

Figure 1
Figure 1. Integration of QML with BCI Systems. II. BACKGROUND A. Fundamentals of Quantum Computing QC utilizes the principles of quantum mechanics to perform computation. A quantum state |ψ⟩ can be defined in a complex vector space known as a Hilbert space H. Quantum bits, or qubits, are the building blocks of quantum information. The ability of qubits to exist in superpositions of states [24] is a key principle [25]. It can … view at source ↗
Figure 2
Figure 2. Quantum Circuit of QNNs Process. Algorithm 1 Quantum Neural Network (QNN) Input: Dataset Output: Evaluation Metrics 1: Import necessary libraries and modules 2: Load and preprocess the dataset 3: function QUANTUMCIRCUIT(parameters, x) 4: Create a quantum circuit 5: Apply quantum gates based on parameters and features 6: Measure the output qubit 7: return qc 8: end function 9: function COSTFUNCTION(parameters, X, y) … view at source ↗
Figure 3
Figure 3. QSVM Quantum Circuit for Classification. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Quantum Circuit of QSVM-QNN Process. QSVM-QNN, one layer of parametrized circuit is used, while 5 epochs of QNN are used to collect the optimal parameters to classify the test data points. For all the classical and quantum algorithms, the test size is taken to be 0.2 f…
Figure 5
Figure 5. Figure 5: Accuracy and Loss for QNN and QSVM-QNN on EEGMM [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Accuracy Comparison across Different Noise Models for the [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 39 canonical work pages

  1. [1]

    Brain computer interfaces, a review,

    L. F. Nicolas-Alonso and J. Gomez-Gil, “Brain computer interfaces, a review,” Sensors, vol. 12, no. 2, pp. 1211–1279, 2012. 12

    work page 2012

  2. [2]

    Brain-computer interface: trend, challenges, and threats,

    B. Maiseli, A. T. Abdalla, L. V . Massawe, M. Mbise, K. Mkocha, N. A. Nassor, M. Ismail, J. Michael, and S. Kimambo, “Brain-computer interface: trend, challenges, and threats,” Brain Informatics , vol. 10, 2023

    work page 2023

  3. [3]

    Combining brain–computer interfaces and assistive technologies: State-of-the-art and challenges,

    J. d. R. Mill ´an, R. Rupp, G. R. M ¨uller-Putz, R. Murray-Smith, C. Giugliemma, M. Tangermann, C. Vidaurre, F. Cincotti, A. K ¨ubler, R. Leeb, C. Neuper, K.-R. M ¨uller, and D. Mattia, “Combining brain–computer interfaces and assistive technologies: State-of-the-art and challenges,” Frontiers in Neuroscience, vol. 4, p. 161, 2010

    work page 2010

  4. [4]

    Brain–machine interfaces: past, present and future,

    M. A. Lebedev and M. A. L. Nicolelis, “Brain–machine interfaces: past, present and future,” Trends in Neurosciences , vol. 29, no. 9, pp. 536– 546, 2006

    work page 2006

  5. [5]

    Brain-computer interfaces for communication and control,

    J. R. Wolpaw, N. Birbaumer, D. J. McFarland, G. Pfurtscheller, and T. M. Vaughan, “Brain-computer interfaces for communication and control,” Clinical Neurophysiology, vol. 113, no. 6, pp. 767–791, 2002

    work page 2002

  6. [6]

    Bci2000: A general-purpose brain-computer interface (bci) system,

    G. Schalk, D. J. McFarland, T. Hinterberger, N. Birbaumer, and J. R. Wolpaw, “Bci2000: A general-purpose brain-computer interface (bci) system,” IEEE Transactions on Biomedical Engineering , vol. 51, no. 6, pp. 1034–1043, 2004

    work page 2004

  7. [7]

    A review of classification algorithms for eeg-based brain–computer interfaces,

    F. Lotte, M. Congedo, A. L ´ecuyer, F. Lamarche, and B. Arnaldi, “A review of classification algorithms for eeg-based brain–computer interfaces,” Journal of Neural Engineering , vol. 4, no. 2, pp. R1–R13, 2007

    work page 2007

  8. [8]

    Machine learning for real-time single-trial eeg- analysis: From brain–computer interfacing to mental state monitoring,

    K.-R. M ¨uller, M. Tangermann, G. Dornhege, M. Krauledat, G. Curio, and B. Blankertz, “Machine learning for real-time single-trial eeg- analysis: From brain–computer interfacing to mental state monitoring,” Journal of Neuroscience Methods , vol. 167, no. 1, pp. 82–90, 2008

    work page 2008

Show all 43 references
  1. [9]

    Deep learning for electroen- cephalogram (eeg) classification tasks: A review,

    A. Craik, Y . He, and J. L. Contreras-Vidal, “Deep learning for electroen- cephalogram (eeg) classification tasks: A review,” Journal of Neural Engineering, vol. 16, no. 3, p. 031001, 2019

    2019

  2. [10]

    A review of classification algorithms for eeg- based brain–computer interfaces: a 10 year update,

    F. Lotte, L. Bougrain, A. Cichocki, M. Clerc, M. Congedo, A. Rako- tomamonjy, and F. Yger, “A review of classification algorithms for eeg- based brain–computer interfaces: a 10 year update,” Journal of Neural Engineering, vol. 15, no. 3, p. 031005, 2018

    2018

  3. [11]

    Deep learning with convolutional neural networks for eeg decoding and visualization,

    R. T. Schirrmeister, J. T. Springenberg, L. D. J. Fiederer, M. Glasstetter, K. Eggensperger, M. Tangermann, F. Hutter, W. Burgard, and T. Ball, “Deep learning with convolutional neural networks for eeg decoding and visualization,” Human Brain Mapping , vol. 38, no. 11, pp. 539...

    2017

  4. [12]

    Deep learning-based electroencephalography analysis: A systematic review,

    Y . Roy, H. Banville, I. Albuquerque, A. Gramfort, T. H. Falk, and J. Faubert, “Deep learning-based electroencephalography analysis: A systematic review,” Journal of Neural Engineering , vol. 16, no. 5, p. 051001, 2019

    2019

  5. [13]

    Transfer learning in brain-computer interfaces,

    V . Jayaram, M. Alamgir, Y . Altun, B. Scholkopf, and M. Grosse- Wentrup, “Transfer learning in brain-computer interfaces,” IEEE Com- putational Intelligence Magazine , vol. 11, no. 1, pp. 20–31, 2016

    2016

  6. [14]

    The cybathlon bci race: Successful longitudinal mutual learning with two tetraplegic users,

    S. Perdikis, L. Tonin, S. Saeedi, C. Schneider, and J. d. R. Mill ´an, “The cybathlon bci race: Successful longitudinal mutual learning with two tetraplegic users,” PLoS Biology, vol. 16, no. 5, p. e2003787, 2018

    2018

  7. [15]

    A multiscale convolutional neural network for eeg-based motor imagery classification,

    M. Zhou, H. He, and Z. Wang, “A multiscale convolutional neural network for eeg-based motor imagery classification,” Biomedical Signal Processing and Control, vol. 74, p. 103101, 2022

    2022

  8. [16]

    Deep fusion of convolutional and re- current neural networks for eeg-based emotion recognition,

    X. Zhao, Y . Zhang, and Z. Li, “Deep fusion of convolutional and re- current neural networks for eeg-based emotion recognition,” Biomedical Signal Processing and Control , vol. 74, p. 103496, 2022

    2022

  9. [17]

    A hybrid attention-based cnn-lstm model for eeg emotion recognition,

    H. Liu, P. Li, and Y . Wang, “A hybrid attention-based cnn-lstm model for eeg emotion recognition,” Engineering Applications of Artificial Intelligence, vol. 113, p. 105347, 2022

    2022

  10. [18]

    A new univariate feature selec- tion algorithm based on the best–worst multi-attribute decision-making method,

    D. P. M. Abellana and D. M. Lao, “A new univariate feature selec- tion algorithm based on the best–worst multi-attribute decision-making method,” Decision Analytics Journal , vol. 7, p. 100240, 2023

    2023

  11. [19]

    Brain computer interfacing: Applications and challenges,

    S. N. Abdulkader, A. Atia, and M. S. Mostafa, “Brain computer interfacing: Applications and challenges,” Egyptian Informatics Journal, vol. 16, no. 2, pp. 213–230, 2015

    2015

  12. [20]

    Schuld and F

    M. Schuld and F. Petruccione, Machine Learning with Quantum Com- puters, 2nd ed. Springer, 2021

    2021

  13. [21]

    Quantum machine learning,

    J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, “Quantum machine learning,” Nature, vol. 549, no. 7671, pp. 195–202, 2017

    2017

  14. [22]

    Quantum machine learning in feature hilbert spaces,

    M. Schuld and N. Killoran, “Quantum machine learning in feature hilbert spaces,” Physical Review Letters , vol. 122, p. 040504, 2019

    2019

  15. [23]

    Quantum autoencoders via quantum adders with genetic algorithms,

    L. Lamata, U. Alvarez-Rodriguez, J. D. Martin-Guerrero, M. Sanz, and E. Solano, “Quantum autoencoders via quantum adders with genetic algorithms,” Quantum Science and Technology, vol. 4, no. 1, p. 014007, 2020

    2020

  16. [24]

    Discussion of probability relations between separated systems,

    E. Schr ¨odinger, “Discussion of probability relations between separated systems,” Mathematical Proceedings of the Cambridge Philosophical Society, vol. 31, no. 4, pp. 555–563, 1935

    1935

  17. [25]

    On the constitution of atoms and molecules,

    N. Bohr, “On the constitution of atoms and molecules,” Philosophical Magazine, vol. 26, no. 151, pp. 1–25, 1913

    1913

  18. [26]

    Can quantum-mechanical de- scription of physical reality be considered complete?

    A. Einstein, B. Podolsky, and N. Rosen, “Can quantum-mechanical de- scription of physical reality be considered complete?” Physical Review, vol. 47, no. 10, pp. 777–780, 1935

    1935

  19. [27]

    Quantum theory, the church-turing principle and the universal quantum computer,

    D. Deutsch, “Quantum theory, the church-turing principle and the universal quantum computer,” Proceedings of the Royal Society of London A, vol. 400, no. 1818, pp. 97–117, 1985

    1985

  20. [28]

    Algorithms for quantum computation: discrete logarithms and factoring,

    P. W. Shor, “Algorithms for quantum computation: discrete logarithms and factoring,” Proceedings 35th Annual Symposium on Foundations of Computer Science, pp. 124–134, 1994

    1994

  21. [29]

    A fast quantum mechanical algorithm for database search,

    L. K. Grover, “A fast quantum mechanical algorithm for database search,” Proceedings of the 28th Annual ACM Symposium on Theory of Computing, pp. 212–219, 1996

    1996

  22. [30]

    Circuit-centric quantum classifiers,

    M. Schuld, A. Bocharov, K. Svore, and N. Wiebe, “Circuit-centric quantum classifiers,” Physical Review A , vol. 101, p. 032308, 2020

    2020

  23. [31]

    Supervised learning with quantum computers,

    M. Schuld and F. Petruccione, “Supervised learning with quantum computers,” Quantum Science and Technology, vol. 3, no. 3, p. 030502, 2018

    2018

  24. [32]

    Quantum algorithms for supervised and unsupervised machine learning,

    S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum algorithms for supervised and unsupervised machine learning,” arXiv preprint arXiv:1307.0411, 2013

    2013 arXiv

  25. [33]

    Quantum-enhanced mea- surements: Beating the standard quantum limit,

    V . Giovannetti, S. Lloyd, and L. Macccone, “Quantum-enhanced mea- surements: Beating the standard quantum limit,” SCIENCE, vol. 306, no. 5700, pp. 1330–1336, 2004

    2004

  26. [34]

    Investigation of quantum support vector machine for classification in nisq era,

    A. Kariya and B. K. Behera, “Investigation of quantum support vector machine for classification in nisq era,” arXiv:2112.06912, 2021

    2021 arXiv

  27. [35]

    Explainable quantum clustering method to model medical data,

    S. Deshmukh, B. K. Behera, P. Mulay, E. A. Ahmed, S. Al-Kuwari, P. Tiwari, and A. Farouk, “Explainable quantum clustering method to model medical data,” Knowledge-Based Systems, vol. 267, 2023

    2023

  28. [36]

    Eeg motor movement/imagery dataset,

    B. Leung, “Eeg motor movement/imagery dataset,” https://www.kaggle. com/code/brianleung2020/eeg-motor-movement, 2022

    2022

  29. [37]

    Eeg dataset,

    S. Nikolas, “Eeg dataset,” https://www.kaggle.com/datasets/samnikolas/ eeg-dataset, 2022

    2022

  30. [38]

    Analysis of quantum machine learning algorithms in noisy channels for classification tasks in the iot extreme environment,

    S. K. Satpathy, V . Vibhu, B. K. Behera, S. Al-Kuwari, S. Mumtaz, and A. Farouk, “Analysis of quantum machine learning algorithms in noisy channels for classification tasks in the iot extreme environment,” IEEE Internet of Things Journal , pp. 1 – 1, 2023

    2023

  31. [39]

    Differential entropy feature for eeg-based emotion classification,

    R.-N. Duan, J.-Y . Zhu, and B.-L. Lu, “Differential entropy feature for eeg-based emotion classification,” in 2013 6th International IEEE/EMBS Conference on Neural Engineering (NER) . IEEE, 2013, pp. 81–84

    2013

  32. [40]

    Eegeyenet: A simultaneous electroencephalog- raphy and eye-tracking dataset and benchmark for eye movement prediction,

    A. Kastrati, M. B. Płomecka, D. Pascual, L. Wolf, V . Gillioz, R. Watten- hofer, and N. Langer, “Eegeyenet: A simultaneous electroencephalog- raphy and eye-tracking dataset and benchmark for eye movement prediction,” arXiv preprint arXiv:2111.05100 , 2021

    2021 arXiv

  33. [41]

    Deap: A database for emotion analysis using physiological signals,

    S. Koelstra, C. Muhl, M. Soleymani, J.-S. Lee, A. Yazdani, T. Ebrahimi, T. Pun, A. Nijholt, and I. Patras, “Deap: A database for emotion analysis using physiological signals,” IEEE Transactions on Affective Computing, vol. 3, no. 1, pp. 18–31, 2012

    2012

  34. [42]

    A large eeg dataset for studying cross-session variability in motor imagery brain-computer interface,

    J. Ma, B. Yang, W. Qiu, Y . Li, S. Gao, and X. Xia, “A large eeg dataset for studying cross-session variability in motor imagery brain-computer interface,” Scientific Data, vol. 9, no. 1, p. 531, 2022

    2022

  35. [43]

    Training embedding quantum kernels with data re-uploading quantum neural networks,

    P. Rodriguez-Grasa, Y . Ban, and M. Sanz, “Training embedding quantum kernels with data re-uploading quantum neural networks,” arXiv:2401.04642, 2024

    2024 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.