CoChat
Sign Up
PreprintCited by 35Quantum Computing Algorithms and ArchitectureQuantum Information and CryptographyQuantum many-body systems

The DOI matches its DataCite record.

The Capacity of Quantum Neural Networks

Open access
Wright, Logan G., McMahon, Peter L.·2019·arXiv
DOI 10.48550/arxiv.1908.01364arXiv 1908.01364
Open this paper in CoChat — free accountOpen PDF

Abstract

A key open question in quantum computation is what advantages quantum neural networks (QNNs) may have over classical neural networks (NNs), and in what situations these advantages may transpire. Here we address this question by studying the memory capacity $C$ of QNNs, which is a metric of the expressive power of a QNN that we have adapted from classical NN theory. We present a capacity inequality showing that the capacity of a QNN is bounded by the information $W$ that can be trained into its parameters: $C \leq W$. One consequence of this bound is that QNNs that are parameterized classically do not show an advantage in capacity over classical NNs having an equal number of parameters. However, QNNs that are parametrized with quantum states could have exponentially larger capacities. We illustrate our theoretical results with numerical experiments by simulating a particular QNN based on a Gaussian Boson Sampler. We also study the influence of sampling due to wavefunction collapse during operation of the QNN, and provide an analytical expression connecting the capacity to the number of times the quantum system is measured.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

DataCite · 10.48550/arxiv.1908.01364

Cites

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • A Gaussian Boson Sampler can be used for quantum reservoir computing to perform machine learning tasks such as classification.

    “A Gaussian Boson Sampler (see Sec. III.B) can also be used for quantum reservoir computing as suggested in (Wright and McMahon, 2019) to perform machine learning tasks such as classification.”

    From Noisy intermediate-scale quantum algorithms (2022) · cites this paper as Wright and McMahon, 2019

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

No references could be resolved yet.

Research works printed but not linked: 70

  1. E. Behrman, J. Niemel, J. Steck, and S. Skinner, in Pro- ceedings of the 4th Workshop on Physics of Computation (1996).
  2. E. Farhi and H. Neven, arXiv:1802.06002. · arXiv
  3. E. Farhi, J. Goldstone, and S. Gutmann, arXiv:1411.4028. · arXiv
  4. A. Peruzzo, J. McClean, P. Shadbolt, M.-H. Yung, X.-Q. Zhou, P. J. Love, A. Aspuru-Guzik, and J. L. O'Brien, Nature Communications 5, 4213 (2014). · DOI · Paper page
  5. J. R. McClean, J. Romero, R. Babbush, and A. Aspuru- Guzik, New Journal of Physics 18, 023023 (2016). · DOI · Paper page
  6. M. Benedetti, E. Lloyd, and S. Sack, arXiv:1906.07682. · arXiv
  7. M. Schuld, I. Sinayskiy, and F. Petruccione, Quantum Information Processing 13, 2567 (2016). · DOI · Paper page
  8. K. H. Wan, O. Dahlsten, H. Kristjánsson, R. Gardner, and M. S. Kim, npj Quantum Information 3, 36 (2017). · DOI · Paper page
  9. P. Rebentrost, T. R. Bromley, C. Weedbrook, and S. Lloyd, Physical Review A 98, 042308 (2018).
  10. N. Killoran, T. R. Bromley, J. M. Arrazola, M. Schuld, N. Quesada, and S. Lloyd, arXiv:1806.06871. · arXiv
  11. G. R. Steinbrecher, J. P. Olson, D. Englund, and J. Car- olan, npj Quantum Information 5, 60 (2019). · DOI
  12. K. Beer, D. Bondarenko, T. Farrelly, T. J. Osborne, R. Salzmann, and R. Wolf, arXiv:1902.10445. · arXiv
  13. F. Tacchino, C. Macchiavello, D. Gerace, and D. Bajoni, npj Quantum Information 5, 26 (2019). · DOI
  14. I. Cong, S. Choi, and M. D. Lukin, arXiv:1810.03787. · arXiv
  15. J. Carolan, M. Mosheni, J. P. Olson, M. Prabhu, C. Chen, D. Bunandar, N. C. Harris, F. N. C. Wong, M. Hochberg, S. Lloyd, and D. Englund, arXiv:1904.10463. · arXiv
  16. N. Wiebe, A. Kapoor, and K. M. Svore, arXiv:1412.3489. · arXiv
  17. M. Benedetti, J. Realpe-Gómez, R. Biswas, and A. Perdomo-Ortiz, Physical Review A 94, 022308 (2016). · DOI · Paper page
  18. G. Verdon, M. Broughton, and J. Biamonte, arXiv:1712.05304. · arXiv
  19. S. H. Adachi and M. P. Henderson, arXiv:1510.06356. · arXiv
  20. M. H. Amin, E. Andriyash, J. Rolfe, B. Kulchytskyy, and R. Melko, Physical Review X 8, 021050 (2018). · DOI · Paper page
  21. K. Mitarai, M. Negoro, M. Kitagawa, and K. Fujii, Phys- ical Review A 98, 032309 (2018). · DOI · Paper page
  22. C. M. Wilson, J. S. Otterbach, N. Tezak, R. S. Smith, G. E. Crooks, and M. P. da Silva, arXiv:1806.08321. · arXiv
  23. M. Schuld, A. Bocharov, K. Svore, and N. Wiebe, arXiv:1804.00633. · arXiv
  24. M. Schuld and N. Killoran, Physical Review Letters 122, 040504 (2019). · DOI · Paper page
  25. V. Havlíček, A. D. Córcoles, K. Temme, A. W. Harrow, A. Kandala, J. M. Chow, and J. M. Gambetta, Nature 567, 209 (2019). · DOI · Paper page
  26. D. Ventura and T. Martinez, arXiv:9807053.
  27. N. Wiebe, A. Kapoor, and K. Svore, arXiv:1401.2142. · arXiv
  28. S. Arunachalam and R. de Wolf, arXiv:1701.06806. · arXiv
  29. G. Verdon, J. Pye, and M. Broughton, arXiv:1806.09729. · arXiv
  30. J. D. Biamonte and P. J. Love, Physical Review A 78, 012352 (2008). · DOI
  31. J. Biamonte, arXiv:1903.04500. · arXiv
  32. J. Preskill, Quantum 2, 79 (2018). · DOI · Paper page
  33. K. M. Nakanishi, K. Fujii, and S. Todo, arXiv:1903.12166. · arXiv
  34. V. Bergholm, J. Izaac, M. Schuld, C. Gogolin, C. Blank, K. McKiernan, and N. Killoran, arXiv:1811.04968. · arXiv
  35. M. Schuld, V. Bergholm, C. Gogolin, J. Izaac, and N. Killoran, Physical Review A 99, 032331 (2019). · DOI · Paper page
  36. J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Babbush, and H. Neven, Nature Communications 9, 4812 (2018). · DOI · Paper page
  37. R. M. Parrish, E. G. Hohenstein, P. L. McMahon, and T. J. Martinez, arXiv:1906.08728. · arXiv
  38. V. N. Vapnik and A. Y. Chervonenkis, in Measures of Complexity (Springer International Publishing, Cham, 2015) pp. 11-30. · DOI
  39. D. J. C. Mackay, Information Theory, Inference, and Learning Algorithms (Learning, 2005). · DOI
  40. N. Tishby and N. Zaslavsky, arXiv:1503.02406. · arXiv
  41. R. Shwartz-Ziv and N. Tishby, arXiv:1703.00810. · arXiv
  42. G. Friedland and M. Krell, arXiv:1708.06019. · arXiv
  43. P. Baldi and R. Vershynin, Advances in Neural Informa- tion Processing Systems, 7729-7738 (2018).
  44. G. Friedland, A. Metere, and M. Krell, arXiv:1810.02328. · arXiv
  45. P. Baldi and R. Vershynin, arXiv:1901.00434. · arXiv
  46. T. M. Cover, IEEE Transactions on Electronic Comput- ers 3, 326 (1965). · DOI · Paper page
  47. E. Gardner, Europhysics Letters 4, 481 (1987). · DOI
  48. W. Kinzel, Philosophical Magazine B 77, 1455 (1998). · DOI
  49. S. Shalev-Shwartz and S. Ben-David, Understanding ma- chine learning: From theory to algorithms (Cambridge University Press, 2013). · DOI · Paper page
  50. V. N. Vapnik, The nature of statistical learning theory (Springer, 2000).
  51. G. B. Huang, Q. Y. Zhu, and C. K. Siew, in IEEE In- ternational Conference on Neural Networks -Conference Proceedings (2004). · DOI
  52. K. Fujii and K. Nakajima, Physical Review Applied 8, 024030 (2017). · DOI · Paper page
  53. M. Negoro, K. Mitarai, K. Fujii, K. Nakajima, and M. Kitagawa, arXiv:1806.10910. · arXiv
  54. K. Nakajima, K. Fujii, M. Negoro, K. Mitarai, and M. Kitagawa, arXiv:1803.04574. · arXiv
  55. A. Kutvonen, T. Sagawa, and K. Fujii, arXiv:1807.03947. · arXiv
  56. S. Ghosh, A. Opala, M. Matuszewski, T. Paterek, and T. C. H. Liew, npj Quantum Information 5, 35 (2019). · DOI · Paper page
  57. J. Chen and H. I. Nurdin, arXiv:1901.01653. · arXiv
  58. H. Jaeger, in GMD-Forschungszentrum Information- stechnik Report 148 (2001). · DOI
  59. W. Maass, T. Natschläger, and H. Markram, Neural Computation 14, 2531 (2002). · DOI · Paper page
  60. S. Aaronson and A. Arkhipov, in Proceedings of the 43rd annual ACM symposium on Theory of computing -STOC '11 (ACM Press, New York, New York, USA, 2011) p. 333. · DOI · Paper page
  61. A. Lund, A. Laing, S. Rahimi-Keshari, T. Rudolph, J. O'Brien, and T. Ralph, Physical Review Letters 113, 100502 (2014). · DOI
  62. C. S. Hamilton, R. Kruse, L. Sansoni, S. Barkhofen, C. Silberhorn, and I. Jex, Physical Review Letters 119, 170501 (2017). · DOI · Paper page
  63. J. Huh, G. G. Guerreschi, B. Peropadre, J. R. McClean, and A. Aspuru-Guzik, Nature Photonics 9, 615 (2015). · DOI · Paper page
  64. K. Brádler, P.-L. Dallaire-Demers, P. Rebentrost, D. Su, and C. Weedbrook, Physical Review A 98, 032310 (2018). · DOI
  65. J. M. Arrazola, T. R. Bromley, and P. Rebentrost, Phys- ical Review A 98, 012322 (2018). · DOI
  66. M. Schuld, K. Brádler, R. Israel, D. Su, and B. Gupt, arXiv:1905.12646. · arXiv
  67. N. Killoran, J. Izaac, N. Quesada, V. Bergholm, M. Amy, and C. Weedbrook, arXiv:1804.03159. · arXiv
  68. J. Johansson, P. Nation, and F. Nori, Computer Physics Communications 184, 1234 (2013). · DOI
  69. L. Banchi and S. Pirandola, in Quantum Information and Measurement (QIM) V: Quantum Technologies (OSA, Washington, D.C., 2019) p. S4B.5. · DOI
  70. E. Dolzhkov, B. Ghandchi, and D. O. Theis, arXiv:1903.12611. The Capacity of Quantum Neural Networks: Supplemental Material Detailed description of the GBS-fQRC Our GBS-fQRC reservoir circuit consists of M input squeezers, followed by a random M -mode interferometer and finally, measurement in the Fock basis. Quantum data is input as states directly into the circuit. We input classical data, x i , to the device by encoding it in the squeezing parameter of the squeezed-vacuum states (i.e., the ith squeezer is realized with an operator Ŝi ∝ exp[x i (â 2 i -(â † i ) . The choice of a nonlinear embedding, rather than state-amplitude encoding, is an important part of transforming the input classical data into the reservoir circuit's high-dimensional Hilbert space. We consider a quantum-classical hybrid approach in which training and the final layer of the QNN is implemented with a classical computer. The output of the GBS-fQRC reservoir circuit is chosen to be a vector · arXiv

Cited by

Sorted by citations

Metadata

Authors
Wright, L. G., & McMahon, P. L.
Affiliations
Cornell University (US)
ORCID
Venue
arXiv
Publisher
arXiv
Published
2019
Pages
JM4G.5-JM4G.5
DOI
10.48550/arxiv.1908.01364
arXiv
1908.01364
OpenAlex
W2965124577
Cited by
35
References
0
Citations by year
2025: 2 · 2024: 5 · 2023: 4 · 2022: 10 · 2021: 11 · 2020: 3
FWCI
2.67
Type
conference paper
Language
en
Open access
Licence
arxiv.org/licenses/nonexclusive-distrib/1.0/
Locations
  • Conference on Lasers and Electro-Optics (publishedVersion)
  • arXiv (Cornell University) (submittedVersion)
  • Figshare (submittedVersion, cc-by)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum Information and Cryptography
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum many-body systems
Keywords
quantum neural networks, expressiveness, information theory, Gaussian boson sampling, quantum machine learning, near-term quantum devices

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review