Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
175 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Learning low-degree quantum objects (2405.10933v1)

Published 17 May 2024 in quant-ph, cs.CC, cs.DS, cs.LG, and math.FA

Abstract: We consider the problem of learning low-degree quantum objects up to $\varepsilon$-error in $\ell_2$-distance. We show the following results: $(i)$ unknown $n$-qubit degree-$d$ (in the Pauli basis) quantum channels and unitaries can be learned using $O(1/\varepsilond)$ queries (independent of $n$), $(ii)$ polynomials $p:{-1,1}n\rightarrow [-1,1]$ arising from $d$-query quantum algorithms can be classically learned from $O((1/\varepsilon)d\cdot \log n)$ many random examples $(x,p(x))$ (which implies learnability even for $d=O(\log n)$), and $(iii)$ degree-$d$ polynomials $p:{-1,1}n\to [-1,1]$ can be learned through $O(1/\varepsilond)$ queries to a quantum unitary $U_p$ that block-encodes $p$. Our main technical contributions are new Bohnenblust-Hille inequalities for quantum channels and completely bounded~polynomials.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (74)
  1. Forrelation: A problem that optimally separates quantum from classical computing. In Proceedings of the Forty-Seventh Annual ACM Symposium on Theory of Computing, STOC ’15, page 307–316, New York, NY, USA, 2015. Association for Computing Machinery. doi:10.1145/2746539.2746547.
  2. Polynomials, quantum query complexity, and Grothendieck’s inequality. In 31st Conference on Computational Complexity, CCC 2016, pages 25:1–25:19, 2016. arXiv:1511.08682.
  3. Optimal Algorithms for Learning Quantum Phase States. In 18th Conference on the Theory of Quantum Computation, Communication and Cryptography (TQC 2023), volume 266 of Leibniz International Proceedings in Informatics (LIPIcs), pages 3:1–3:24, 2023.
  4. Quantum query algorithms are completely bounded forms. SIAM J. Comput, 48(3):903–925, 2019. Preliminary version in ITCS’18.
  5. Improved bounds on Fourier entropy and min-entropy. ACM Transactions on Computation Theory (TOCT), 13(4):1–40, 2021.
  6. Two new results about quantum exact learning. Quantum, 5:587, 2021.
  7. Approximating the cut-norm via Grothendieck’s inequality. In Proceedings of the Thirty-Sixth Annual ACM Symposium on Theory of Computing, STOC ’04, page 72–80, New York, NY, USA, 2004. Association for Computing Machinery. doi:10.1145/1007352.1007371.
  8. Quantum algorithms for learning and testing juntas. Quantum Information Processing, 6(5):323–348, 2007.
  9. Quantum lower bounds by polynomials. J. ACM, 48(4):778–797, 2001.
  10. On the absolute convergence of Dirichlet series. Annals of Mathematics, pages 600–622, 1931.
  11. Learning DNF over the uniform distribution using a quantum example oracle. In Proceedings of the Eighth Annual Conference on Computational Learning Theory, COLT ’95, page 118–127, New York, NY, USA, 1995. Association for Computing Machinery. doi:10.1145/225298.225312.
  12. The Grothendieck constant is strictly smaller than Krivine’s bound. Forum Math. Pi, 1:453–462, 2013. Preliminary version in FOCS’11. arXiv:1103.6161.
  13. Harald Bohr. Ueber die Bedeutung der Potenzreihen unendlich vieler Variablen in der Theorie der Dirichletschen Reihen. Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen, Mathematisch-Physikalische Klasse, 1913:441–488, 1913.
  14. The Bohr radius of the n𝑛nitalic_n-dimensional polydisk is equivalent to (log⁡n)/n𝑛𝑛(\log n)/n( roman_log italic_n ) / italic_n. Advances in Mathematics, 264:726–746, 2014.
  15. k-forrelation optimally separates quantum and classical query complexity. In Proceedings of the 53rd Annual ACM SIGACT Symposium on Theory of Computing, STOC 2021, page 1303–1316, New York, NY, USA, 2021. Association for Computing Machinery. doi:10.1145/3406325.3451040.
  16. Influence in Completely Bounded Block-Multilinear Forms and Classical Simulation of Quantum Algorithms. In 37th Computational Complexity Conference (CCC 2022), volume 234, pages 28:1–28:21, 2022. doi:10.4230/LIPIcs.CCC.2022.28.
  17. Guest column: Approximate degree in classical and quantum computing. ACM SIGACT News, 51(4):48–72, 2021.
  18. On Testing and Learning Quantum Junta Channels. In Gergely Neu and Lorenzo Rosasco, editors, Proceedings of Thirty Sixth Conference on Learning Theory, volume 195 of Proceedings of Machine Learning Research, pages 1064–1094. PMLR, 12–15 Jul 2023. URL: https://proceedings.mlr.press/v195/bao23b.html.
  19. Clément L Canonne. A short note on learning discrete distributions. arXiv preprint arXiv:2002.11457, 2020.
  20. The Power of Block-Encoded Matrix Powers: Improved Regression Techniques via Faster Hamiltonian Simulation. In Christel Baier, Ioannis Chatzigiannakis, Paola Flocchini, and Stefano Leonardi, editors, 46th International Colloquium on Automata, Languages, and Programming (ICALP 2019), volume 132 of Leibniz International Proceedings in Informatics (LIPIcs), pages 33:1–33:14, Dagstuhl, Germany, 2019. Schloss Dagstuhl – Leibniz-Zentrum für Informatik. URL: https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ICALP.2019.33, doi:10.4230/LIPIcs.ICALP.2019.33.
  21. Sample efficient algorithms for learning quantum channels in pac model and the approximate state discrimination problem. arXiv preprint arXiv:1810.10938, 2018.
  22. Speed limits and locality in many-body quantum dynamics. Reports on Progress in Physics, 2023.
  23. Testing and Learning Quantum Juntas Nearly Optimally. In Proceedings of the 2023 Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pages 1163–1185. SIAM, 2023.
  24. Geometry of multilinear forms. Communications in Contemporary Mathematics, 22(02):1950011, 2020.
  25. A. Davie. Lower bound for KGsubscript𝐾𝐺K_{G}italic_K start_POSTSUBSCRIPT italic_G end_POSTSUBSCRIPT. Unpublished, 1984.
  26. AM Davie. Matrix norms related to Grothendieck’s inequality. In Banach Spaces: Proceedings of the Missouri Conference held in Columbia, USA, June 24–29, 1984, pages 22–26. Springer, 2006.
  27. The Bohnenblust–Hille inequality for homogeneous polynomials is hypercontractive. Annals of mathematics, pages 485–497, 2011.
  28. Lower bounds for the constants in the Bohnenblust–Hille inequality: The case of real scalars. Proceedings of the American Mathematical Society, 142(2):575–580, 2014.
  29. On the Fourier spectrum of functions on boolean cubes. Mathematische Annalen, 374:653–680, 2019.
  30. Learning low-degree functions from a logarithmic number of random queries. In Proceedings of the 54th Annual ACM SIGACT Symposium on Theory of Computing, STOC 2022, page 203–207, New York, NY, USA, 2022. Association for Computing Machinery. doi:10.1145/3519935.3519981.
  31. Low-degree learning and the metric entropy of polynomials. arXiv preprint arXiv:2203.09659, 2022.
  32. Francisco Escudero Gutiérrez. Influences of Fourier completely bounded polynomials and classical simulation of quantum algorithms. arXiv:2304.06713, 2023.
  33. Pauli error estimation via population recovery. Quantum, 5:549, 2021.
  34. Lower bounds on learning Pauli channels. arXiv preprint arXiv:2301.09192, 2023.
  35. Efficient estimation of Pauli channels. ACM Transactions on Quantum Computing, 1(1), dec 2020. doi:10.1145/3408039.
  36. Process tomography for unitary quantum channels. Journal of Mathematical Physics, 55(3), 2014.
  37. Alexandre Grothendieck. Résumé de la théorie métrique des produits tensoriels topologiques. Soc. de Matemática de São Paulo, 1953.
  38. Uffe Haagerup. A new upper bound for the complex Grothendieck constant. Israel Journal of Mathematics, 60:199–224, 1987.
  39. Learning to predict arbitrary quantum processes. PRX Quantum, 4:040337, Dec 2023. URL: https://link.aps.org/doi/10.1103/PRXQuantum.4.040337, doi:10.1103/PRXQuantum.4.040337.
  40. Sample-optimal tomography of quantum states. IEEE Transactions on Information Theory, 63(9):5628–5641, 2017.
  41. Predicting many properties of a quantum system from very few measurements. Nature Physics, 16(10):1050–1057, 2020.
  42. Fast estimation of sparse quantum noise. PRX Quantum, 2:010322, Feb 2021.
  43. Tight bounds on the Fourier growth of bounded functions on the hypercube. arXiv preprint arXiv:2107.06309, 2021.
  44. Quantum merlin-arthur proof systems: Are multiple merlins more helpful to arthur? In Algorithms and Computation: 14th International Symposium, ISAAC 2003, pages 189–198. Springer, 2003.
  45. Quantum and Classical Low-Degree Learning via a Dimension-Free Remez Inequality. In 15th Innovations in Theoretical Computer Science Conference (ITCS 2024), volume 287 of Leibniz International Proceedings in Informatics (LIPIcs), pages 69:1–69:22, 2024.
  46. Learning quantum circuits of T-depth one. In 2022 IEEE International Symposium on Information Theory (ISIT), pages 2213–2218, 2022. doi:10.1109/ISIT50566.2022.9834452.
  47. John E Littlewood. On bounded bilinear forms in an infinite number of variables. The Quarterly Journal of Mathematics, pages 164–174, 1930.
  48. Constant depth circuits, Fourier transform, and learnability. Journal of the ACM (JACM), 40(3):607–620, 1993.
  49. Richard A. Low. Learning and testing algorithms for the clifford group. Phys. Rev. A, 80:052314, Nov 2009.
  50. A geometric technique to generate lower estimates for the constants in the bohnenblust–hille inequalities. arXiv preprint arXiv:1203.0793, 2012.
  51. Characterizing quantum gates via randomized benchmarking. Phys. Rev. A, 85:042311, Apr 2012. URL: https://link.aps.org/doi/10.1103/PhysRevA.85.042311, doi:10.1103/PhysRevA.85.042311.
  52. Quantum Boolean functions. arXiv preprint arXiv:0810.2435, 2008.
  53. Ashley Montanaro. Some applications of hypercontractive inequalities in quantum information theory. Journal of Mathematical Physics, 53(12), 2012.
  54. There exist multilinear Bohnenblust–Hille constants cnsubscript𝑐𝑛c_{n}italic_c start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT with lim(cn+1−cn)subscript𝑐𝑛1subscript𝑐𝑛\lim(c_{n+1}-c_{n})roman_lim ( italic_c start_POSTSUBSCRIPT italic_n + 1 end_POSTSUBSCRIPT - italic_c start_POSTSUBSCRIPT italic_n end_POSTSUBSCRIPT )= 0. Journal of Functional Analysis, 264(2):429–463, 2013.
  55. Quantum Computation and Quantum Information: 10th Anniversary Edition. Cambridge University Press, Cambridge, UK, 2010. doi:10.1017/CBO9780511976667.
  56. On the Pauli spectrum of QAC0. arXiv preprint arXiv:2311.09631, 2023.
  57. Ryan O’Donnell. Analysis of Boolean Functions. Cambridge University Press, 2014.
  58. Efficient quantum tomography. In Proceedings of the Forty-Eighth Annual ACM Symposium on Theory of Computing, STOC ’16, page 899–912, New York, NY, USA, 2016. Association for Computing Machinery. doi:10.1145/2897518.2897544.
  59. Vern Paulsen. Completely Bounded Maps and Operator Algebras. Cambridge Studies in Advanced Mathematics. Cambridge University Press, 02 2003.
  60. Unbounded violation of tripartite Bell inequalities. Communications in Mathematical Physics, 279:455–486, 2008.
  61. Gilles Pisier. Grothendieck’s theorem, past and present. Bulletin of the American Mathematical Society, 49(2):237–323, 2012.
  62. New upper bounds for the constants in the Bohnenblust–Hille inequality. Journal of Mathematical Analysis and Applications, 386(1):300–307, 2012.
  63. Towards sharp Bohnenblust–Hille constants. Communications in Contemporary Mathematics, 20(03):1750029, 2018.
  64. J. Reeds. A new lower bound on the real Grothendieck constant. Manuscript (http://www.dtc.umn.edu/~reedsj/bound2.dvi), 1991.
  65. Bohnenblust–Hille inequality for cyclic groups. arXiv preprint arXiv:2305.10560, 2023.
  66. Noncommutative bohnenblust-hille inequality in the heisenberg-weyl and gell-mann bases with applications to fast learning. arXiv preprint arXiv:2301.01438, 2023.
  67. Barbara M. Terhal. Quantum error correction for quantum memories. Rev. Mod. Phys., 87:307–346, Apr 2015.
  68. B. S. Tsirelson. Quantum generalizations of Bell’s inequality. Letters in Mathematical Physics, 1980. doi:10.1007/BF00417500.
  69. Locality and error mitigation of quantum circuits. arXiv preprint arXiv:2303.06496, 2023.
  70. Leslie G. Valiant. A Theory of the Learnable. Communications of the ACM, 27(11):1134–1142, 1984.
  71. N. Th. Varopoulos. On an inequality of von Neumann and an application of the metric theory of tensor products to operators theory. J. Functional Analysis, 16:83–100, 1974. doi:10.1016/0022-1236(74)90071-8.
  72. Fernando Vieira Costa Júnior. The optimal multilinear Bohnenblust–Hille constants: A computational solution for the real case. Numerical Functional Analysis and Optimization, 39(15):1656–1668, 2018.
  73. Probabilistic error cancellation with sparse Pauli–Lindblad models on noisy quantum processors. Nature Physics, pages 1–6, 2023.
  74. Noncommutative Bohnenblust–Hille inequalities. Mathematische Annalen, pages 1–20, 2023.
Citations (4)

Summary

We haven't generated a summary for this paper yet.

X Twitter Logo Streamline Icon: https://streamlinehq.com