Convergence Rates of Tseng's Splitting Method and Its Acceleration Schemes for Monotone Inclusion Problem with a Sum of H\"older Continuous Operators
Pith reviewed 2026-06-26 10:22 UTC · model grok-4.3
The pith
Tseng's splitting method and its accelerations converge at rates set by the Hölder exponent for monotone inclusions with summed Hölder continuous operators.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
For the monotone inclusion problem formed as the sum of two monotone operators that are each Hölder continuous with exponent in (0,1], Tseng's splitting method converges at a rate governed by that exponent. The composite extra anchored gradient method and the symplectic composite extra gradient method improve the rate in a manner consistent with the same exponent.
What carries the argument
Tseng's splitting method extended to Hölder continuous monotone operators, which produces convergence rates controlled by the Hölder exponent.
Load-bearing premise
The two operators in the inclusion problem are monotone and Hölder continuous with some fixed exponent in (0,1].
What would settle it
A concrete monotone inclusion problem whose operators are Hölder continuous but for which Tseng's method or its accelerations fail to meet the predicted rate would disprove the claim.
read the original abstract
The monotone inclusion problem is fundamental in applied mathematics and is closely related to a wide range of practical applications. However, existing solution methods typically require the underlying operator to be Lipschitz continuous. Recently, H\"older continuity, a weaker condition than Lipschitz continuity, has proven useful in characterizing certain real-world problems. To bridge this gap, we investigate the convergence rates of the Tseng's splitting method (a fundamental algorithm for monotone inclusion problem) and its two accelerated variants, the composite extra anchored gradient method and the symplectic composite extra gradient method, under the H\"older continuity assumption. Our numerical experiments demonstrate that the numerical performance of these algorithms aligns with their respective theoretical convergence rates.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims to derive convergence rates (determined by the Hölder exponent) for Tseng's splitting method and its two accelerated variants—the composite extra anchored gradient method and the symplectic composite extra gradient method—applied to the monotone inclusion problem 0 ∈ A(x) + B(x) where both operators are monotone and Hölder continuous (with exponent in (0,1]), and reports that numerical experiments align with the theoretical rates.
Significance. If the rates are correctly derived under the stated assumptions, the work would extend splitting-method theory from the standard Lipschitz setting to the weaker Hölder-continuous case, which is relevant for applications where Lipschitz continuity fails. The presence of numerical validation is a strength.
major comments (1)
- [Problem statement / abstract and §1] Problem statement / abstract and §1: the setting is described only as a 'sum of Hölder Continuous Operators' without identifying which operator is maximal monotone. Tseng's method (and its accelerations) requires the resolvent (I + λA)^{-1} of one maximal monotone operator; without this explicit structural split the algorithm is not well-defined and the claimed rates cannot hold.
Simulated Author's Rebuttal
We thank the referee for the thorough review and constructive comments. We address the major comment point by point below.
read point-by-point responses
-
Referee: [Problem statement / abstract and §1] Problem statement / abstract and §1: the setting is described only as a 'sum of Hölder Continuous Operators' without identifying which operator is maximal monotone. Tseng's method (and its accelerations) requires the resolvent (I + λA)^{-1} of one maximal monotone operator; without this explicit structural split the algorithm is not well-defined and the claimed rates cannot hold.
Authors: We agree with the referee that the problem statement requires clarification on the structural assumptions. In the manuscript, we consider the inclusion 0 ∈ A(x) + B(x), where A is maximal monotone and B is monotone and Hölder continuous. However, we acknowledge that this distinction was not explicitly highlighted in the abstract and introduction. We will revise the abstract, §1, and the problem formulation section to clearly specify that A is maximal monotone (to ensure the resolvent is defined) and B is monotone and Hölder continuous with exponent in (0,1]. This revision will make the setting unambiguous without changing the results or proofs. revision: yes
Circularity Check
No circularity: rates derived from stated monotonicity + Hölder assumptions
full rationale
The paper states the monotone inclusion problem with Hölder continuous operators as the setting and derives convergence rates for Tseng's method and accelerations from those assumptions. No self-definitional reduction, fitted-input-as-prediction, or load-bearing self-citation is present in the abstract or described claims. The derivation chain remains independent of the target rates; the result is not forced by definition or prior author work.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption The operators are monotone and Hölder continuous with some exponent alpha in (0,1].
Reference graph
Works this paper leans on
-
[1]
D. G. Anderson. Iterative procedures for nonlinear integ ral equations. Journal of the ACM (JACM) , 12(4): 547–560, 1965. doi: 10.1145/321296.321305
-
[2]
A. S. Antipin. On a method for convex programs using a symme trical modification of the Lagrange function. Economika i Matem. Metody , (6):1164–1173, 1976
1976
-
[3]
Ben-Tal, L
A. Ben-Tal, L. E. Ghaoui, and A. Nemirovski. Robust optimization, volume 28. Princeton university press, Princeton, 2009
2009
-
[4]
R. I. Bot ¸, E. R. Csetnek, and D.-K. Nguyen. Fast optimisti c gradient descent ascent (OGDA) method in continuous and discrete time. F oundations of Computational Mathematics , pages 1–60, 2023. doi: https: //doi.org/10.1007/s10208-023-09636-5
-
[5]
G. Cai, Q.-L. Dong, and Y . Peng. Strong convergence theore ms for solving variational inequality problems with pseudo-monotone and non-lipschitz operators. Journal of Optimization Theory and Applications, 188(2): 447–472, 2021. doi: 10.1007/s10957-020-01792-w
-
[6]
Y . Cai, A. Oikonomou, and W. Zheng. Accelerated algorithm s for constrained nonconvex-nonconcave min-max optimization and comonotone inclusion. In R. Salak hutdinov, Z. Kolter, K. Heller, A. Weller, N. Oliver, J. Scarlett, and F. Berkenkamp, editors, Proceedings of the 41st International Conference on Machine Learning , volume 235 of Proceedings of Machi...
2024
-
[7]
A. Chakraborty and A. Nedi´ c. Popov mirror-prox method fo r variational inequalities, 2025. URL https://arxiv.org/abs/2507.23395
arXiv 2025
-
[8]
K. Chen, D. Sun, Y . Y uan, G. Zhang, and X. Zhao. HPR-LP: An im plementation of an HPR method for solving linear programming. Mathematical Programming Computation , pages 1–28, 2025. doi: 10.1007/ s12532-025-00292-0
2025
-
[9]
C. D. Dang and G. Lan. On the convergence properties of non- euclidean extragradient methods for variational inequalities with generalized monotone operators. Computational Optimization and applications , 60(2):277– 310, 2015. doi: 10.1007/s10589-014-9673-9
-
[10]
C. Daskalakis and I. Panageas. The limit points of (optim istic) gradient descent in min-max optimization. In Proceedings of the 32nd International Conference on Neural Information Processing Systems, NIPS’18, page 9256–9266, 2018. URL https://dl.acm.org/doi/pdf/10.5555/3327546.3327597
-
[11]
Daskalakis, A
C. Daskalakis, A. Ilyas, V . Syrgkanis, and H. Zeng. Train ing GANs with optimism. In International Conference on Learning Representations, 2018. URL https://openreview.net/forum?id=SJJySbbAZ
2018
-
[12]
Diakonikolas, C
J. Diakonikolas, C. Daskalakis, and M. I. Jordan. Efficie nt methods for structured nonconvex-nonconcave min-max optimization. In International Conference on Artificial Intelligence and St atistics, pages 2746–2754. PMLR, 2021
2021
-
[13]
Y . Fan, Y . Li, and B. Chen. Weaker MVI condition: Extragra dient methods with multi-step exploration. In The Twelfth International Conference on Learning Represen tations, 2023
2023
-
[14]
Gorbunov, H
E. Gorbunov, H. Berard, G. Gidel, and N. Loizou. Stochast ic extragradient: General analysis and improved rates. In International Conference on Artificial Intelligence and St atistics, pages 7865–7901. PMLR, 2022
2022
-
[15]
B. Halpern. Fixed points of nonexpanding maps. Bulletin of the American Mathematical Society, 73:957–961,
-
[16]
26 YI ZHANG
URL https://api.semanticscholar.org/CorpusID:120539954. 26 YI ZHANG
-
[17]
M. Ito, Z. Lu, and C. He. A parameter-free conditional gra dient method for composite minimization under H¨ older condition. Journal of Machine Learning Research , 24(166):1–34, 2023. URL http://jmlr.org/papers/v24/22-0983.html
2023
-
[18]
A. N. Iusem, A. Jofr´ e, R. I. Oliveira, and P . Thompson. Ex tragradient method with variance reduction for stochastic variational inequalities. SIAM Journal on Optimization , 27(2):686–724, 2017. doi: 10.1137/ 15M1031953
2017
-
[19]
A. Kannan and U. V . Shanbhag. Optimal stochastic extragr adient schemes for pseudomonotone stochastic variational inequality problems and their variants. Computational Optimization and Applications , 74(3):779– 820, 2019. doi: 10.1007/s10589-019-00120-x
-
[20]
Kinderlehrer and G
D. Kinderlehrer and G. Stampacchia. An introduction to variational inequalities and their applications. SIAM, 2000
2000
-
[21]
G. M. Korpelevich. The extragradient method for finding s addle points and other problems. Matecon, 12: 747–756, 1976
1976
-
[22]
Lee and D
S. Lee and D. Kim. Fast extra gradient methods for smooth s tructured nonconvex-nonconcave minimax problems. In A. Beygelzimer, Y . Dauphin, P . Liang, and J. W. V aughan, editors, Advances in Neural Information Processing Systems, 2021
2021
-
[23]
F. Lieder. On the convergence rate of the Halpern-iterat ion. Optimization letters, 15(2):405–418, 2021. doi: 10.1007/s11590-020-01617-9
- [24]
-
[25]
G. J. Minty. Monotone (nonlinear) operators in Hilbert s pace. Duke Mathematical Journal, 29(3):341 – 346,
-
[26]
URL https://doi.org/10.1215/S0012-7094-62-02933-2
doi: 10.1215/S0012-7094-62-02933-2. URL https://doi.org/10.1215/S0012-7094-62-02933-2
-
[27]
Mishchenko, D
K. Mishchenko, D. Kovalev, E. Shulgin, P . Richt´ arik, and Y . Malitsky. Revisiting stochastic extragradient. In International Conference on Artificial Intelligence and St atistics, pages 4573–4582. PMLR, 2020
2020
-
[28]
A. Nemirovski. Prox-method with rate of convergence o(1 /t) for variational inequalities with lipschitz continuous monotone operators and smooth convex-concave s addle point problems. SIAM Journal on Optimization, 15(1):229–251, 2004. doi: 10.1137/S1052623403425629
-
[29]
Nesterov
Y . Nesterov. A method of solving a convex programming pro blem with convergence rate o(1/k2). Doklady Akademii Nauk SSSR, 269(3):543, 1983
1983
-
[30]
Y . Ouyang and Y . Xu. Lower complexity bounds of first-orde r methods for convex-concave bilinear saddle- point problems. Mathematical Programming, 185(1):1–35, 2021. doi: 10.1007/s10107-019-01420-0
-
[31]
Pethick, P
T. Pethick, P . Latafat, P . Patrinos, O. Fercoq, and V . Cev her. Escaping limit cycles: Global convergence for constrained nonconvex-nonconcave minimax problems. I n International Conference on Learning Representations, 2022. URL https://openreview.net/forum?id=2 vhkAMARk
2022
-
[32]
L. D. Povov. A modification of the Arrow-Hurwitz method of search for saddle points. Mat. Zametki, 28(5): 777–784, 1980
1980
-
[33]
H. Qi and H.-K. Xu. Convergence of Halpern’s iteration me thod with applications in optimization. Numerical Functional Analysis and Optimization, 42(15):1839–1854, 2021. doi: https://doi.org/10.1080/ 01630563.2021. 2001826
arXiv 2021
-
[34]
X. Qu, W. Bian, and X. Chen. An extra gradient Anderson-ac celerated algorithm for pseudomonotone variational inequalities. Mathematics of Computation, 95(359):1361–1387, 2026. doi: 10.1090/mcom/4095
-
[35]
Sedlmayer, D.-K
M. Sedlmayer, D.-K. Nguyen, and R. I. Bot ¸. A fast optimis tic method for monotone variational inequalities. In Proceedings of the 40th International Conference on Machin e Learning, ICML’23, 2023
2023
-
[36]
F. Stonyakin, A. Gasnikov, P . Dvurechensky, A. Titov, an d M. Alkousa. Generalized mirror prox algorithm for monotone variational inequalities: Universality and i nexact oracle. Journal of Optimization Theory and Applications, 194(3):988–1013, 2022. doi: 10.1007/s10957-022-02062- 7
-
[37]
D. Sun, Y . Y uan, G. Zhang, and X. Zhao. Accelerating preco nditioned ADMM via degenerate proximal point mappings. SIAM Journal on Optimization, 35(2):1165–1193, 2025. doi: 10.1137/24M1650053
-
[38]
J. E. Taylor and M. P . Bendsøe. An interpretation for min- max structural design problems including a method for relaxing constraints. International Journal of Solids and Structures , 20(4):301–314, 1984. doi: 10.1016/ ALGORITHMS FOR MONOTONE INCLUSION PROBLEM WITH SUMS OF H ¨OLDER OPERA TORS 27 0020-7683(84)90041-6
1984
-
[39]
Toyasaki, P
F. Toyasaki, P . Daniele, and T. Wakolbinger. A variation al inequality formulation of equilibrium models for end-of-life products with nonlinear constraints. European Journal of Operational Research, 236(1):340–350,
-
[40]
doi: 10.1016/j.ejor.2013.12.006
- [41]
-
[42]
Q. Tran-Dinh. Extragradient-type methods with O(1/k) last-iterate convergence rates for co-hypomonotone inclusions. Journal of Global Optimization , 89(1):197–221, 2024. doi: 10.1007/s10898-023-01347-z
-
[43]
Q. Tran-Dinh and N. Nguyen-Trung. Accelerated extragra dient-type methods – part 2: Generalization and sublinear convergence rates under co-hypomonotonicity, 2 025. URL https://arxiv.org/abs/2501.04585
-
[44]
P . Tseng. A modified forward-backward splitting method f or maximal monotone mappings. SIAM Journal on Control and Optimization, 38(2):431–446, 2000. doi: 10.1137/S0363012998338806
-
[45]
T. Ui. Bayesian Nash equilibrium and variational inequa lities. Journal of Mathematical Economics , 63: 139–146, 2016. ISSN 0304-4068. doi: 10.1016/j.jmateco.20 16.02.004
-
[46]
Y . xiang Y uan and Y . Zhang. Symplectic extra-gradient ty pe method for solving general non-monotone inclusion problem, 2025. URL https://arxiv.org/abs/2406.10793
arXiv 2025
-
[47]
Y oon and E
T. Y oon and E. K. Ryu. Accelerated algorithms for smooth c onvex-concave minimax problems with O(1/k2) rate on squared gradient norm. In International Conference on Machine Learning , pages 12098–12109. PMLR, 2021
2021
-
[48]
Y .-x. Y uan and Y . Zhang. Symplectic discretization appr oach for developing new proximal point algorithm. Computational Optimization and Applications , 93(2):651–687, 2026. doi: 10.1007/s10589-025-00728-2
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.