pith. machine review for the scientific record. sign in

arxiv: 2605.10299 · v1 · submitted 2026-05-11 · 💻 cs.LG

Recognition: no theorem link

Nearly-Optimal Algorithm for Adversarial Kernelized Bandits

Authors on Pith no claims yet

Pith reviewed 2026-05-12 04:11 UTC · model grok-4.3

classification 💻 cs.LG
keywords kernelized banditsadversarial banditsexponential weightsinformation gainregret boundsgaussian process banditsNyström approximation
0
0 comments X

The pith

The exponential-weight algorithm achieves Õ(√(T γ_T)) adversarial regret in kernelized bandits.

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

This paper studies kernelized bandits in an adversarial setting, where an adversary chooses reward functions each round from a known reproducing kernel Hilbert space. It shows that the exponential-weight algorithm attains a regret of Õ(√(T γ_T)), with T the total rounds and γ_T the maximum information gain of the kernel. For squared exponential and ν-Matérn kernels, matching lower bounds prove that the bound is optimal up to polylogarithmic factors. The paper also supplies a computationally efficient version that uses Nyström approximation while keeping the same regret guarantee.

Core claim

The exponential-weight algorithm achieves Õ(√(T γ_T)) adversarial regret in the kernelized bandit problem, where γ_T denotes the maximum information gain. Algorithm-independent lower bounds for the squared exponential and ν-Matérn kernels establish that this performance is optimal up to polylogarithmic factors. A Nyström-based variant preserves the regret bound while remaining computationally tractable.

What carries the argument

The exponential-weight algorithm that maintains a multiplicative distribution over actions based on observed rewards, together with the maximum information gain γ_T that quantifies the complexity of functions in the reproducing kernel Hilbert space.

Load-bearing premise

Reward functions chosen adversarially at each round belong to a fixed known reproducing kernel Hilbert space whose maximum information gain can be bounded.

What would settle it

An explicit adversary and kernel for which every algorithm must incur regret larger than Õ(√(T γ_T)) polylog factors would disprove the upper bound.

read the original abstract

This paper studies kernelized bandits (also known as Gaussian process bandits) in an adversarial environment, where the reward functions in a known reproducing kernel Hilbert space (RKHS) may be adversarially chosen at each round. We show that the exponential-weight algorithm achieves $\tilde{O}(\sqrt{T \gamma_T})$ adversarial regret, where $T$ and $\gamma_T$ denote the number of total rounds and the maximum information gain, respectively. For squared exponential (SE) and $\nu$-Mat\'ern kernels, we also show algorithm-independent lower bounds that guarantee the optimality of our algorithm up to polylogarithmic factors. Furthermore, we present a computationally efficient variant of our algorithm using Nystr\"om approximation while maintaining nearly optimal regret guarantees.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper studies adversarial kernelized bandits where reward functions lie in a known RKHS with fixed kernel. It shows that the exponential-weights algorithm attains Õ(√(T γ_T)) regret, where γ_T is the maximum information gain. Algorithm-independent lower bounds are given for squared-exponential and ν-Matérn kernels that match up to polylog factors, establishing near-optimality. A Nyström-approximated computationally efficient variant is also presented that preserves the nearly-optimal regret.

Significance. If the central regret bounds and lower-bound constructions hold, the work provides the first nearly-tight characterization of adversarial regret for kernelized bandits, directly extending the information-gain framework from the stochastic setting. The explicit dependence on γ_T yields kernel-specific rates (including polylog optimality for SE and Matérn), and the Nyström variant addresses computational tractability without sacrificing the leading term. These are standard strengths of the kernel-bandit literature and are cleanly realized here.

minor comments (3)
  1. [Abstract] Abstract and introduction: the phrase 'nearly optimal' is used for both the main algorithm and the Nyström variant; a brief sentence clarifying that the polylog gap is the only sub-optimality would improve precision.
  2. [Problem formulation] The problem setup should explicitly restate the bounded RKHS-norm assumption (||f||_H ≤ B) alongside the known-kernel assumption, as this is load-bearing for the regret analysis even though it is standard.
  3. [Preliminaries] Notation: the definition of the maximum information gain γ_T should be recalled in the main text (not only in the abstract) before the regret statement, to aid readers unfamiliar with the kernel literature.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive review and recommendation of minor revision. The report correctly summarizes our main results on the exponential-weights algorithm achieving Õ(√(T γ_T)) adversarial regret for kernelized bandits, the algorithm-independent lower bounds matching up to polylog factors for squared-exponential and ν-Matérn kernels, and the Nyström-based computationally efficient variant that preserves the leading regret term.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's central result applies the exponential-weights algorithm to an implicit expert set whose cardinality is controlled by the maximum information gain γ_T of a fixed kernel. This γ_T is an external, kernel-dependent quantity defined independently of the algorithm (standard in the RKHS bandit literature) and is not fitted or derived within the paper. The Õ(√(T γ_T)) regret bound follows from the standard analysis of exponential weights once the RKHS ball and bounded-norm assumptions are granted in the problem setup. The algorithm-independent lower bounds for SE and Matérn kernels rely on known, externally established rates for γ_T and do not reduce to any fitted parameter or self-referential definition. No load-bearing self-citation chains, ansatz smuggling, or renaming of known results as new derivations appear in the derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the standard assumption that rewards belong to a fixed known RKHS; no free parameters, invented entities, or non-standard axioms are introduced in the abstract.

axioms (1)
  • domain assumption Reward functions lie in a known reproducing kernel Hilbert space with bounded maximum information gain γ_T
    Stated in abstract as the setting for the adversarial environment.

pith-pipeline@v0.9.0 · 5414 in / 1221 out tokens · 29173 ms · 2026-05-12T04:11:06.189852+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

82 extracted references · 82 canonical work pages

  1. [1]

    J. D. Abernethy, E. Hazan, and A. Rakhlin. Interior-point methods for full-information and bandit online learning.IEEE Transactions on Information Theory, 58(7):4164–4175, 2012

  2. [2]

    Allenberg, P

    C. Allenberg, P. Auer, L. Gy¨orfi, and G. Ottucs´ak. Hannan consistency in on-line learning in case of unbounded losses under partial monitoring. InInternational Conference on Algorithmic Learning Theory, pages 229–243. Springer, 2006

  3. [3]

    Bartlett, V

    P. Bartlett, V. Dani, T. Hayes, S. Kakade, A. Rakhlin, and A. Tewari. High-probability regret bounds for bandit online linear optimization. InProceedings of the 21st annual conference on learning theory-COLT 2008, 2008

  4. [4]

    Bogunovic, J

    I. Bogunovic, J. Scarlett, S. Jegelka, and V. Cevher. Adversarially robust optimization with Gaussian processes. InProc. Neural Information Processing Systems (NeurIPS), 2018

  5. [5]

    Bogunovic, A

    I. Bogunovic, A. Krause, and J. Scarlett. Corruption-tolerant Gaussian process bandit optimiza- tion. InInternational Conference on Artificial Intelligence and Statistics, 2020

  6. [6]

    Bogunovic, Z

    I. Bogunovic, Z. Li, A. Krause, and J. Scarlett. A robust phased elimination algorithm for corruption-tolerant Gaussian process bandits. InProc. Neural Information Processing Systems (NeurIPS), 2022

  7. [7]

    Bubeck and A

    S. Bubeck and A. Slivkins. The best of both worlds: Stochastic and adversarial bandits. In Conference on Learning Theory, pages 42–1. JMLR Workshop and Conference Proceedings, 2012

  8. [8]

    Bubeck, M

    S. Bubeck, M. Cohen, and Y. Li. Sparsity, variance and curvature in multi-armed bandits. In Algorithmic Learning Theory, pages 111–127. PMLR, 2018

  9. [9]

    A. D. Bull. Convergence rates of efficient global optimization algorithms.Journal of Machine Learning Research, 2011

  10. [10]

    Cai and J

    X. Cai and J. Scarlett. On lower bounds for standard and robust Gaussian process bandit optimization. InProceedings of the 38th International Conference on Machine Learning, volume 139 ofProceedings of Machine Learning Research, pages 1216–1226. PMLR, 2021

  11. [11]

    Cai and J

    X. Cai and J. Scarlett. Lower bounds for time-varying kernelized bandits. InInternational Conference on Artificial Intelligence and Statistics, pages 73–81. PMLR, 2025

  12. [12]

    X. Cai, S. Gomes, and J. Scarlett. Lenient regret and good-action identification in Gaussian process bandits. InInternational Conference on Machine Learning, pages 1183–1192. PMLR, 2021

  13. [13]

    Calandriello and L

    D. Calandriello and L. Rosasco. Statistical and computational trade-offs in kernel k-means. Advances in neural information processing systems, 31, 2018

  14. [14]

    Calandriello, A

    D. Calandriello, A. Lazaric, and M. Valko. Distributed adaptive sampling for kernel matrix approximation. InArtificial Intelligence and Statistics, 2017

  15. [15]

    Calandriello, L

    D. Calandriello, L. Carratino, A. Lazaric, M. Valko, and L. Rosasco. Gaussian process optimization with adaptive sketching: Scalable and no regret. InConference on learning theory, 2019

  16. [16]

    Camilleri, K

    R. Camilleri, K. Jamieson, and J. Katz-Samuels. High-dimensional experimental design and kernel bandits. InProc. International Conference on Machine Learning (ICML), 2021

  17. [17]

    Chatterji, A

    N. Chatterji, A. Pacchiano, and P. Bartlett. Online learning with kernel losses. InInternational Conference on Machine Learning, 2019. 10

  18. [18]

    S. R. Chowdhury and A. Gopalan. On kernelized multi-armed bandits. InProc. International Conference on Machine Learning (ICML), 2017

  19. [19]

    S. R. Chowdhury and A. Gopalan. No-regret algorithms for multi-task Bayesian optimization. InInternational Conference on Artificial Intelligence and Statistics, 2021

  20. [20]

    Y. Deng, X. Zhou, B. Kim, A. Tewari, A. Gupta, and N. Shroff. Weighted Gaussian process bandits for non-stationary environments. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2022

  21. [21]

    Desautels, A

    T. Desautels, A. Krause, and J. W. Burdick. Parallelizing exploration-exploitation tradeoffs in Gaussian process bandit optimization.Journal of Machine Learning Research, 2014

  22. [22]

    V. V. Fedorov.Theory of optimal experiments. Elsevier, 2013

  23. [23]

    Han and J

    E. Han and J. Scarlett. Adversarial attacks on Gaussian process bandits. InInternational Conference on Machine Learning, 2022

  24. [24]

    Hazan and S

    E. Hazan and S. Kale. A simple multi-armed bandit algorithm with optimal variation-bounded regret. InProceedings of the 24th Annual Conference on Learning Theory, pages 817–820. JMLR Workshop and Conference Proceedings, 2011

  25. [25]

    K. Hong, Y. Li, and A. Tewari. An optimization-based algorithm for non-stationary kernel bandits without prior knowledge. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2023

  26. [26]

    S. Iwazaki. Gaussian process upper confidence bound achieves nearly-optimal regret in noise- free Gaussian process bandits. InThe Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

  27. [27]

    S. Iwazaki. Improved regret bounds for Gaussian process upper confidence bound in Bayesian optimization. InThe Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025

  28. [28]

    S. Iwazaki. Tighter regret lower bound for Gaussian process bandits with squared exponential kernel in hypersphere.arXiv preprint arXiv:2602.17940, 2026

  29. [29]

    Iwazaki and S

    S. Iwazaki and S. Takeno. Improved regret analysis in Gaussian process bandits: Optimality for noiseless reward, RKHS norm, and non-stationary variance. InInternational Conference on Machine Learning, pages 26642–26672. PMLR, 2025

  30. [30]

    Iwazaki and S

    S. Iwazaki and S. Takeno. Near-optimal algorithm for non-stationary kernelized bandits. In International Conference on Artificial Intelligence and Statistics, pages 406–414. PMLR, 2025

  31. [31]

    Iwazaki, Y

    S. Iwazaki, Y. Inatsu, and I. Takeuchi. Mean-variance analysis in Bayesian optimization under uncertainty. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2021

  32. [32]

    Janz.Sequential decision making with feature-linear models

    D. Janz.Sequential decision making with feature-linear models. PhD thesis, 2022

  33. [33]

    Gaussian Processes and Kernel Methods: A Review on Connections and Equivalences

    M. Kanagawa, P. Hennig, D. Sejdinovic, and B. K. Sriperumbudur. Gaussian processes and kernel methods: A review on connections and equivalences.arXiv preprint arXiv:1807.02582, 2018

  34. [34]

    Kirschner, I

    J. Kirschner, I. Bogunovic, S. Jegelka, and A. Krause. Distributionally robust Bayesian optimization. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2020

  35. [35]

    Krause and C

    A. Krause and C. Ong. Contextual Gaussian process bandit optimization. InProc. Neural Information Processing Systems (NeurIPS), 2011

  36. [36]

    Lattimore and C

    T. Lattimore and C. Szepesv ´ari.Bandit algorithms. Cambridge University Press, 2020

  37. [37]

    C.-W. Lee, H. Luo, C.-Y. Wei, M. Zhang, and X. Zhang. Achieving near instance-optimality and minimax-optimality in stochastic and adversarial linear bandits simultaneously. InInternational Conference on Machine Learning, 2021

  38. [38]

    Li and J

    Z. Li and J. Scarlett. Gaussian process bandit optimization with few batches. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2022

  39. [39]

    Lizotte, T

    D. Lizotte, T. Wang, M. Bowling, and D. Schuurmans. Automatic gait optimization with Gaussian process regression. InProc. International Joint Conference on Artificial Intelligence (IJCAI), 2007. 11

  40. [40]

    Mason, L

    B. Mason, L. Jain, S. Mukherjee, R. Camilleri, K. Jamieson, and R. Nowak. Nearly optimal algorithms for level set estimation. InInternational Conference on Artificial Intelligence and Statistics, 2022

  41. [41]

    Musco and C

    C. Musco and C. Musco. Recursive sampling for the Nystr ¨om method.Advances in neural information processing systems, 30, 2017

  42. [42]

    G. Neu, J. Olkhovskaya, and S. Vakili. Adversarial contextual bandits go kernelized. In International Conference on Algorithmic Learning Theory, 2024

  43. [43]

    Q. P. Nguyen, Z. Dai, B. K. H. Low, and P. Jaillet. Value-at-risk optimization with Gaussian processes. InProc. International Conference on Machine Learning (ICML), 2021

  44. [44]

    C. E. Rasmussen and C. K. I. Williams.Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning). The MIT Press, 2005

  45. [45]

    Riutort-Mayol, P.-C

    G. Riutort-Mayol, P.-C. B¨ urkner, M. R. Andersen, A. Solin, and A. Vehtari. Practical hilbert space approximate Bayesian Gaussian processes for probabilistic programming.Statistics and Computing, 33(1):17, 2023

  46. [46]

    Russo and B

    D. Russo and B. Van Roy. Learning to optimize via posterior sampling.Mathematics of Operations Research, 39(4):1221–1243, 2014

  47. [47]

    Russo and B

    D. Russo and B. Van Roy. Learning to optimize via information-directed sampling.Advances in neural information processing systems, 27, 2014

  48. [48]

    Saday, Y

    A. Saday, Y. C. Yıldırım, and C. Tekin. Robust Bayesian satisficing.Advances in Neural Information Processing Systems, 2023

  49. [49]

    Salgia, S

    S. Salgia, S. Vakili, and Q. Zhao. A domain-shrinking based Bayesian optimization algorithm with order-optimal regret performance. InAdvances in Neural Information Processing Systems, volume 34, pages 28836–28847. Curran Associates, Inc., 2021

  50. [50]

    Salgia, S

    S. Salgia, S. Vakili, and Q. Zhao. Random exploration in Bayesian optimization: Order-optimal regret and computational efficiency. InProc. International Conference on Machine Learning (ICML), 2024

  51. [51]

    Scarlett

    J. Scarlett. Tight regret bounds for Bayesian optimization in one dimension. InProceedings of the 35th International Conference on Machine Learning, volume 80 ofProceedings of Machine Learning Research, pages 4500–4508. PMLR, 2018

  52. [52]

    Scarlett, I

    J. Scarlett, I. Bogunovic, and V. Cevher. Lower bounds on regret for noisy Gaussian process bandit optimization. InProc. Conference on Learning Theory (COLT), 2017

  53. [53]

    P. G. Sessa, I. Bogunovic, M. Kamgarpour, and A. Krause. No-regret learning in unknown games with correlated payoffs.Advances in Neural Information Processing Systems, 32, 2019

  54. [54]

    P. G. Sessa, I. Bogunovic, M. Kamgarpour, and A. Krause. Learning to play sequential games versus unknown opponents.Advances in neural information processing systems, 33:8971–8981, 2020

  55. [55]

    O. Shamir. On the complexity of bandit linear optimization. InConference on Learning Theory, 2015

  56. [56]

    Shekhar and T

    S. Shekhar and T. Javidi. Multi-scale zero-order optimization of smooth functions in an RKHS. arXiv preprint arXiv:2005.04832, 2020

  57. [57]

    Shekhar and T

    S. Shekhar and T. Javidi. Instance dependent regret analysis of kernelized bandits. InInternational Conference on Machine Learning, 2022

  58. [58]

    Snoek, H

    J. Snoek, H. Larochelle, and R. P. Adams. Practical Bayesian optimization of machine learning algorithms. InProc. Neural Information Processing Systems (NeurIPS), 2012

  59. [59]

    Solin and S

    A. Solin and S. S¨arkk¨a. Hilbert space methods for reduced-rank gaussian process regression. Statistics and Computing, 2020

  60. [60]

    Srinivas, A

    N. Srinivas, A. Krause, S. Kakade, and M. Seeger. Gaussian process optimization in the bandit setting: No regret and experimental design. InProc. International Conference on Machine Learning (ICML), 2010

  61. [61]

    Takemori and M

    S. Takemori and M. Sato. Approximation theory based methods for RKHS bandits. In International Conference on Machine Learning, 2021. 12

  62. [62]

    Takeno and S

    S. Takeno and S. Iwazaki. On regret bounds of thompson sampling for bayesian optimization. arXiv preprint arXiv:2603.09276, 2026

  63. [63]

    Takeno, Y

    S. Takeno, Y. Inatsu, and M. Karasuyama. Randomized Gaussian process upper confidence bound with tighter Bayesian regret bounds. InProceedings of the 40th International Conference on Machine Learning, volume 202 ofProceedings of Machine Learning Research, pages 33490–33515. PMLR, 2023

  64. [64]

    S. Vakili. Open problem: Regret bounds for noise-free kernel-based bandits. InProc. Conference on Learning Theory (COLT), 2022

  65. [65]

    Vakili, N

    S. Vakili, N. Bouziani, S. Jalali, A. Bernacchia, and D. shan Shiu. Optimal order simple regret for Gaussian process bandits. InProc. Neural Information Processing Systems (NeurIPS), 2021

  66. [66]

    Vakili, K

    S. Vakili, K. Khezeli, and V. Picheny. On information gain and regret bounds in Gaussian process bandits. InProc. International Conference on Artificial Intelligence and Statistics (AISTATS), 2021

  67. [67]

    Valko, N

    M. Valko, N. Korda, R. Munos, I. Flaounas, and N. Cristianini. Finite-time analysis of kernelised contextual bandits. InProceedings of the Twenty-Ninth Conference on Uncertainty in Artificial Intelligence, UAI’13, page 654–663. AUAI Press, 2013

  68. [68]

    Vershynin.High-dimensional probability: An introduction with applications in data science

    R. Vershynin.High-dimensional probability: An introduction with applications in data science. Cambridge university press, 2018

  69. [69]

    Yadav, D

    M. Yadav, D. Sheldon, and C. Musco. Gaussian process bandits for top-k recommendations. Advances in Neural Information Processing Systems, 2024

  70. [70]

    Zhou and N

    X. Zhou and N. Shroff. No-regret algorithms for time-varying Bayesian optimization. In2021 55th Annual Conference on Information Sciences and Systems (CISS). IEEE, 2021

  71. [71]

    Zimmert and T

    J. Zimmert and T. Lattimore. Return of the bias: Almost minimax optimal high probability bounds for adversarial linear bandits. InConference on Learning Theory, 2022

  72. [72]

    𝑇∑︁ 𝑡=1 𝜎2 (x (𝑇+1) ;X (𝑇) , 𝜆) # (26) ≤E X(𝑇) ,x (𝑇+1)

    M. Zuluaga, A. Krause, et al. e-pal: An active learning approach to the multi-objective optimization problem.Journal of Machine Learning Research, 2016. 13 A Notation table Tab. 2 summarizes the notations used in this paper. B Pseudocode In this section, we provide the pseudocode for the algorithms that could not be included in the main text due to space ...

  73. [73]

    The number𝑠 𝑡 satisfies𝑠 𝑡 =𝑂(𝛾 𝑇 log(𝑇 𝛾𝑇 ))=𝑂(𝛾 𝑇 log(𝑇))for any𝑡∈ [𝑇] 11

  74. [74]

    𝑓𝑡 (x∗) − ∑︁ x∈ X e𝑃𝑡 (x)𝑓 𝑡 (x) # (90) =E [ 𝑓𝑡 (x∗) −e𝑢𝑡 (x∗)]| {z } (i) +E

    The computational cost of recursive RLS-Nystr ¨om is 𝑂(|X|𝛾 2 𝑇 (log(𝑇 𝛾𝑇 )) 2)= 𝑂(|X|𝛾 2 𝑇 (log𝑇) 2). Here, we confirm the computational cost of ( e𝑓𝑡 (x)) x∈ X and (e𝜎2 𝑡 (x)) x∈ X is 𝑂(|X|𝑠 2 𝑡 +𝑠 3 𝑡 ). Firstly, the computation of Nystr ¨om embedding 𝜙𝑡 (·) requires the calculation of pseudo-inverse [(S ⊤ 𝑡 e𝐾𝑡S𝑡 )1/2]†, which require 𝑂(𝑠 3 𝑡 )-comput...

  75. [75]

    Finally, Lemma H.11 is used to obtain the upper bound on the regret arising from the unfavorable rare event, where the accuracy of recursive RLS-Nystr¨om is not sufficient

    and [15], which study the kernel approximation for k-means clustering and the computationally efficient KB algorithm, respectively. Finally, Lemma H.11 is used to obtain the upper bound on the regret arising from the unfavorable rare event, where the accuracy of recursive RLS-Nystr¨om is not sufficient. The proof of Lemma H.11 is obtained with elementary ...

  76. [76]

    The number of sampled columns 𝑠∈ [|X|] satisfies 𝑠≤𝑠 max(4𝛾𝑇 , 𝛿), where 𝑠max (𝑤, 𝑧)= 384(𝑤+1)log((𝑤+1)/𝑧)

  77. [77]

    3 is at most𝑂(|X|𝑠 2 max (4𝛾𝑇 , 𝛿))

    The computational time of Alg. 3 is at most𝑂(|X|𝑠 2 max (4𝛾𝑇 , 𝛿))

  78. [78]

    3 satisfies the following: 1 2 ¯𝚿 ¯𝚿⊤ + 𝜆 𝑇 I| X | ⪯ ¯𝚿SS⊤ ¯𝚿⊤ + 𝜆 𝑇 I| X | ⪯ 3 2 ¯𝚿 ¯𝚿⊤ + 𝜆 𝑇 I| X | ,(235) where ¯𝚿=( ¯𝜓(x (1) ),

    The weighed sampling matrixS∈R | X | ×𝑠 returned by Alg. 3 satisfies the following: 1 2 ¯𝚿 ¯𝚿⊤ + 𝜆 𝑇 I| X | ⪯ ¯𝚿SS⊤ ¯𝚿⊤ + 𝜆 𝑇 I| X | ⪯ 3 2 ¯𝚿 ¯𝚿⊤ + 𝜆 𝑇 I| X | ,(235) where ¯𝚿=( ¯𝜓(x (1) ), . . . , ¯𝜓(x ( | X | ))) ∈R | X | × | X |. 36 In statements 1 and 2, 𝛾𝑇 :=𝛾 𝑇 (𝜆,X) denotes the maximum information gain of kernel 𝑘 on X with the variance parameter𝜆. ...

  79. [79]

    The number of sampled columns e𝑠∈ [𝑇] satisfies e𝑠≤𝑠 max (𝛾𝑇 , 𝛿), where 𝑠max (𝑤, 𝑧)= 384(𝑤+1)log((𝑤+1)/𝑧)

  80. [80]

    3 is at most𝑂(𝑇 𝑠 2 max (𝛾𝑇 , 𝛿))

    The computational time of Alg. 3 is at most𝑂(𝑇 𝑠 2 max (𝛾𝑇 , 𝛿))

Showing first 80 references.