Unifying restart accelerated gradient and proximal bundle methods

Jiaming Liang

REVIEW 2 major objections 5 minor 38 references

Unifying restart accelerated gradient and proximal bundle methods

T0 review · 2 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read A restarted accelerated gradient method and the proximal bundle method are shown to be instances of (accelerated) inexact proximal point frameworks, giving restart ACG optimal complexity.

desk verdict Liang's paper is mostly right: the inner-loop verification is solid, but the main complexity bound is imported without proof and the novelty vs. his earlier work is under-specified. read the letter →

arxiv 2501.04165 v1 pith:MQQP3DN5 submitted 2025-01-07 math.OC

Jiaming Liang This is my paper · ORCID

classification math.OC MSC 49M3765K0568Q2590C2590C3090C60

keywords convexcompositeoptimizationrestartacceleratedgradientproximalbundlemethodhybridextragradientinexactpointiterationcomplexitycutting-planesmooth

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

This paper tries to show that two apparently unrelated optimization methods—the restart accelerated composite gradient (ACG) method and the modern proximal bundle method—are in fact two faces of the same proximal-point principle. For restart ACG, the paper constructs a specific algorithm and proves that each outer loop is precisely an instance of the accelerated hybrid proximal extragradient (A-HPE) framework, which yields the optimal iteration complexity $O(\sqrt{L}d_0/\sqrt{\bar\varepsilon})$ for convex smooth composite problems. For the proximal bundle method, it proves the analogous membership in the non-accelerated HPE framework, showing the bundle method is a restarted cutting-plane method. The value of the claim is explanatory: if true, the practical edge of restarted and bundle schemes over their single-step counterparts is inherited from a common multi-step versus single-step implementation distinction.

What carries the argument

The load-bearing object is the A-HPE framework, an accelerated hybrid proximal extragradient scheme: an outer loop that approximately solves a proximal subproblem, with relative-error conditions relating a subgradient $u_k\in\partial_{\eta_k}\varphi(\tilde{w}_k)$ to the displacement $\tilde{w}_k-\tilde{z}_k$, and then updates along $u_k$. The inner ACG variant is the engine that produces those triples; its key identity converts the algorithm's defining inequalities into exactly the A-HPE conditions. For the bundle method, the corresponding machinery is the HPE framework (the non-accelerated version) and a bundle model $\Gamma_j\le f$; the paper shows the bundle stopping criterion is equivalent to HPE's relative-error inequality. The constant $0.9$ in the restart ACG stopping criterion is arbitrary in $(0,1)$, showing the result is insensitive to that tolerance.

What would settle it

Run Algorithm 2 on a one-dimensional convex composite instance such as $f(x)=x^2$ and $h(x)=|x|$, with $\lambda$ set at both endpoints $1/L$ and $d_0^2/\bar\varepsilon$, and count all inner ACG iterations required to reach objective gap $\bar\varepsilon$ from $x_0=d_0$; if the count grows faster than a constant times $\sqrt{L}d_0/\sqrt{\bar\varepsilon}$, the claimed bound is false.

Watch

Extended reading notes

Core claim

The paper's central claim is that restart ACG is a concrete realization of the A-HPE framework: each call to the inner ACG subroutine produces a triple $(\tilde{w}_k,u_k,\eta_k)$ satisfying the framework's relative-error conditions, and the outer acceleration schedule is exactly A-HPE with constant step size. Relying on the existing A-HPE convergence theorem, the paper concludes that restart ACG solves any convex smooth composite problem in $O(\sqrt{L}d_0/\sqrt{\bar\varepsilon})$ iterations—the same optimal rate as the classical accelerated gradient method. It further proves that the modern proximal bundle method, when its bundle subproblems are solved to a declared tolerance, satisfies the HPE framework's conditions with the assignment $w_k=x_j$, $\tilde{w}_k=\tilde{x}_j$, $u_k=(x_0-x_j)/\lambda$, and $\eta_k=\varphi(\tilde{x}_j)-(\Gamma_j+h)(x_j)+\langle x_0-x_j,x_j-\tilde{x}_j\rangle/\lambda$, making it an HPE instance and hence a restarted cutting-plane method. The paper's stated aim is the unification: both methods are multi-step implementations of (accelerated) inexact proximal point schemes.

Load-bearing premise

The optimal-complexity bound rests entirely on an existing A-HPE theorem that is invoked without proof and assumed to survive the generalization from a smooth function to the composite objective $\varphi$.

Editorial extensions

If this is right

If the main theorem holds, restart ACG is optimal for convex smooth composite problems, matching the classical accelerated gradient rate $O(\sqrt{L}d_0/\sqrt{\bar\varepsilon})$ while suppressing oscillatory behavior.
Every outer iteration needs only a subroutine satisfying the A-HPE relative-error conditions, so any other inner method with the same guarantee can substitute for ACG without changing the outer rate.
The proximal bundle result implies the bundle method is a restarted cutting-plane method, so its optimal $O(M^2d_0^2/\bar\varepsilon^2)$ bound for nonsmooth composite problems follows from HPE theory rather than a separate bundle analysis.
The multi-step versus single-step comparison gives a structural reason why restart ACG and the proximal bundle method outperform their single-step counterparts in practice: both allow larger prox steps and pay for them with inner loops.
If the imported convergence theorem extends exactly as assumed, the proof route is unified: analyze the outer proximal scheme once, then supply any subroutine satisfying the error conditions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The paper leaves implicit that any restart schedule for ACG can be read as a choice of how many inner iterations to run before the next prox step; tuning that schedule is equivalent to tuning the HPE relative-error tolerance.
The same lens suggests a testable design: replace the inner ACG subroutine with a different fast proximal subroutine, such as a second-order or randomized method, and the optimal outer rate should survive as long as the relative-error conditions hold.
For strongly convex objectives with unknown strong-convexity parameter, the framework points toward a universal restart ACG whose step size $\lambda$ is adapted online; whether that achieves optimal complexity without knowing the parameter is open.
Editorial: the unification is structural rather than black-box—it identifies matching stopping criteria and update rules, so it may extend to non-Euclidean or stochastic settings where both A-HPE and HPE have analogues.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's my take on Liang's paper. The cleanest part is the internal math: Proposition 2.8, which shows the inner ACG loop produces an A-HPE triple in O(min{sqrt(lambda*L), ...}) iterations, is verifiable and correct. I checked Lemmas 2.6 and 2.7; the arguments go through, and the constant 9/16 is consistent. The MPB-as-HPE lemma (Lemma 3.1) is a nice, short observation and genuinely clarifies the connection between bundle methods and proximal point schemes.

The weak spot is exactly what the stress-test note says: Theorem 2.5 is the load-bearing complexity estimate, and it is imported from [28] with one sentence about a 'minor modification.' The paper doesn't prove it, and it doesn't identify which sequence in [28] maps to w_k here. That said, I think the concern is more about presentation than substance: Algorithm 2 is literally A-HPE with constant stepsize, and the standard A-HPE theorem does give phi(w_k) for this update rule. Still, a referee should ask the author to prove the theorem in this context or cite a precise statement in [28] that covers composite phi.

My bigger hesitation is novelty. The paper calls Algorithm 2 a 'novel restarted version,' but it is a doubly accelerated inexact proximal point method, which is the title of the author's earlier paper [19]. The paper never clarifies how the restart version differs from [19] in construction or result. That is a real positioning problem.

Overall, the mathematical core is sound, the MPB connection is new, and the complexity bound is likely correct. It deserves a serious referee. I'd send it out with a request: prove or precisely localize Theorem 2.5, and explain the relationship to [19] and [23]. If those are resolved, it is a solid contribution for the proximal point and bundle method community.

Referee Report

2 major / 5 minor

Summary. This paper proposes a restarted accelerated composite gradient (restart ACG) method for convex smooth composite optimization (CSCO) and claims it achieves the optimal iteration complexity O(√L d0/√ε). The method is a double-loop algorithm: the outer loop follows the accelerated hybrid proximal extragradient (A-HPE) framework of Monteiro and Svaiter [28] with a constant proximal step λ, and the inner loop invokes an ACG subroutine to approximately solve the proximal subproblem until the relative-error conditions (13)–(14) are met. Section 2.3 proves that the inner ACG loop produces such a triple in O(min{√(λL), ln(λL)}) iterations, and Theorem 2.9 combines this with an imported rate bound from [28] to obtain the stated optimal complexity. Section 3 shows that the modern proximal bundle (MPB) method is an instance of the HPE framework, giving explicit expressions for the triple satisfying (25)–(26). The paper's stated contribution is a conceptual unification of restart ACG and MPB as multi-step implementations of proximal-point-based acceleration frameworks.

Significance. If the main claim holds, the paper offers a clean unifying perspective: restart ACG is a multi-step implementation of A-HPE and MPB is a multi-step implementation of HPE, which would help explain their practical advantage over single-step counterparts. The inner-loop verification in Proposition 2.8 is concrete and the arithmetic checks out, and Lemma 3.1 is a clean verification that MPB fits the HPE framework. However, the optimal-complexity theorem rests on an unproved import from [28], so the central claim is not yet fully supported in the manuscript. The conceptual contribution is real but conditional on closing that gap.

major comments (2)

[Section 2.2, Theorem 2.5] The paper's central optimal-complexity claim, Theorem 2.9, rests entirely on Theorem 2.5, which is not proved. The paragraph before Theorem 2.5 asserts that 'with minor modification, such as generalizing f to φ, the results in Section 3 of [28] are applicable' and that 'Theorem 3.8 of [28] also holds,' but it does not verify the hypotheses of [28, Thm. 3.8] for Algorithm 2, nor does it identify which of the sequences generated by Algorithm 2 (w_k, \tilde w_k, or z_k) the theorem's bound attaches to. Since [28, Thm. 3.8] is proved for a specific A-HPE sequence and under specific step-size/error-tolerance conditions, the one-sentence 'minor modification' assertion is insufficient. Please provide a proof of Theorem 2.5 or a detailed mapping of Algorithm 2's iterates to the framework of [28, Section 3], and verify explicitly that all assumptions of [28, Thm. 3.8] hold for constant λ and the 0.9-relative-error criterion (14).
[Section 2.2, Algorithm 2 and Theorem 2.5] The bound φ(w_k) − φ* ≤ 2 d0^2/(λ k^2) in Theorem 2.5 is stated 'for completeness' but no proof or derivation is given, and the text explicitly says 'without providing a proof.' This is a load-bearing missing element, not a mere presentation issue. Even if the transfer from [28] is straightforward, the authors should either include the adaptation proof in the paper or state precisely which theorem in [28] applies to which sequence and why the sequence w_k produced by the min-selection in step 3 of Algorithm 2 inherits the bound.

minor comments (5)

[Sections 2.2 and 2.3 (d0)] The quantity d0 appears in Theorem 2.5, Theorem 2.9, and the abstract but is never defined. Please define it explicitly (presumably d0 = ||w0 − z*|| for some minimizer z* of φ).
[Section 2.3, Lemma 2.4] Lemma 2.4 is imported from [27, Prop. 1(c)] without proof and is used to guarantee A_j ≥ 6λ in Proposition 2.8. Since this lemma is load-bearing for the inner-loop complexity, please either include a proof in the appendix or state explicitly which result in [27] is being used and confirm that its hypotheses match the ACG variant in Section 2.1, including the strong-convexity and smoothness constants.
[Section 3.2, complexity claim] The statement that 'MPB and the subgradient method both have optimal complexity bound O(M^2 d0^2/ε^2)' is made without a proof or citation. If this is intended as context, a reference would be helpful; if it is a new claim, it needs support.
[Abstract and Introduction] There are minor typographical issues, e.g., 'acc elerated' and 'smoo th' in the abstract, and 'fconcludes' in the introduction. These should be corrected during revision.
[Section 2.2, Theorem 2.5 labeling] Since Theorem 2.5 is imported rather than proved, consider relabeling it as a quoted result or a lemma with a clear citation, so that the reader can distinguish the paper's original contributions from the imported rate bound.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: restart ACG's optimal bound is an unproved-but-independent import from [28], not an equivalence to its inputs; MPB's HPE-instance is proved directly.

full rationale

The paper's central complexity claim (Theorem 2.9) is not circular. The outer-loop rate in Theorem 2.5 is imported from the independent A-HPE theorem of Monteiro and Svaiter [28], and Proposition 2.8 independently verifies that the inner ACG loop produces triples satisfying (13)-(14). The paper does not prove Theorem 2.5; it states it without proof via a 'minor modification' assertion. That is a support and verifiability gap, not a circularity, because [28] is an external theorem and does not assume the present paper's conclusion. The MPB connection in Lemma 3.1 is proved directly by constructing HPE triples from the bundle optimality conditions, so it does not reduce to its inputs. Self-citations such as [18], [21], and [23] provide the prior MPB method and interpretation; none supplies the HPE-instance proof or the optimal-complexity bound. No equation is defined in terms of the quantity being predicted, and no fitted parameter is relabeled as a prediction. Overall, the derivation chain is self-contained in the sense required for a circularity finding, with any weakness lying in the unproved transfer of an external theorem rather than in a definitional or self-citational circle.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No data fitting occurs; lambda and 0.9 are user-chosen algorithmic constants, not fitted parameters. The central claims rest on two imported technical results, Theorem 3.8 of [28] and Lemma 2.4 of [27], whose exact applicability is assumed rather than proven. No new mathematical entities are postulated.

assumptions (4)

domain assumption Theorem 3.8 of Monteiro-Svaiter [28] applies after 'generalizing f to phi' in the present setting.
The outer-loop complexity bound, Theorem 2.5, is imported rather than derived; if the extension is not valid, the O(sqrt(L/eps)d0) rate is unsupported.
standard math Lemma 2.4, the lower bound A_j at least max{j^2/(4L), (1/L)(1+sqrt(mu)/(2sqrt(L)))^(2(j-1))}, holds for the ACG variant as stated.
The inner-loop iteration count in Proposition 2.8 depends on this bound, whose proof is omitted and delegated to [27].
domain assumption The composite objective phi = f + h satisfies the standard CSCO assumptions: f convex L-smooth, h convex with tractable prox, and dom h subset of dom f.
This is the problem class in (1) and is used in every proximal subproblem.
domain assumption For MPB, the bundle model Gamma_j satisfies Gamma_j <= f and the cutting-plane subproblems (28) have solutions.
Lemma 3.1 needs Gamma_j <= f to establish the epsilon-subgradient inclusion.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Unifying restart accelerated gradient and proximal bundle methods." pith.science (2026). https://pith.science/paper/MQQP3DN5

@misc{pith2026250104165,
  author       = {Pith},
  title        = {Pith review of: Unifying restart accelerated gradient and proximal bundle methods},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MQQP3DN5}},
  note         = {Machine review of arXiv:2501.04165}
}

read the original abstract

This paper presents a novel restarted version of Nesterov's accelerated gradient method and establishes its optimal iteration-complexity for solving convex smooth composite optimization problems. The proposed restart accelerated gradient method is shown to be a specific instance of the accelerated inexact proximal point framework introduced in "An accelerated hybrid proximal extragradient method for convex optimization and its implications to second-order methods" by Monteiro and Svaiter, SIAM Journal on Optimization, 2013. Furthermore, this work examines the proximal bundle method within the inexact proximal point framework, demonstrating that it is an instance of the framework. Notably, this paper provides new insights into the underlying algorithmic principle that unifies two seemingly disparate optimization methods, namely, the restart accelerated gradient and the proximal bundle methods.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 35 canonical work pages

[28]

R. D. C. Monteiro and B. F. Svaiter. An accelerated hybri d proximal extragradient method for convex optimization and its implications to second-ord er methods. SIAM Journal on Optimization, 23(2):1092–1125, 2013

work page 2013
[19]

A Doubly Accelerated Inexact Proximal Point Method for Nonconvex Composite Optimization Problems

J. Liang and R. D. C. Monteiro. A doubly accelerated inex act proximal point method for nonconvex composite optimization problems. Available on arXiv:1811.11378 , 2018

work page Pith review arXiv 2018
[23]

Liang and R

J. Liang and R. D. C. Monteiro. A uniﬁed analysis of a clas s of proximal bundle methods for solving hybrid convex composite optimization problems. Mathematics of Operations Research, 49(2):832–855, 2024

work page 2024
[1]

Ahn and S

K. Ahn and S. Sra. Understanding nesterov’s acceleratio n via proximal point method. In Sym- posium on Simplicity in Algorithms (SOSA) . Society for Industrial and Applied Mathematics, 2022

work page 2022
[2]

Alamo, P

T. Alamo, P. Krupa, and D. Limon. Gradient based restart F ISTA. In 58th IEEE Conference on Decision and Control (CDC) , pages 3936–3941. IEEE, 2019

work page 2019
[3]

Alamo, P

T. Alamo, P. Krupa, and D. Limon. Restart FISTA with globa l linear convergence. In 18th IEEE European Control Conference (ECC) , pages 1969–1974. IEEE, 2019

work page 1969
[4]

Alamo, P Krupa, and D

T. Alamo, P Krupa, and D. Limon. Restart of accelerated ﬁr st order methods with linear convergence under a quadratic functional growth condition . IEEE Transactions on Automatic Control, 67(10):5200–5214, 2022

work page 2022
[5]

Beck and M

A. Beck and M. Teboulle. A fast iterative shrinkage-thre sholding algorithm for linear inverse problems. SIAM Journal on Imaging Sciences , 2(1):183–202, 2009

work page 2009

Show all 38 references

[6]

Bubeck, Y

S. Bubeck, Y. T. Lee, and M. Singh. A geometric alternativ e to nesterov’s accelerated gradient descent. arXiv preprint arXiv:1506.08187 , 2015

2015 arXiv
[7]

Carmon, J

Y. Carmon, J. C. Duchi, O. Hinder, and A. Sidford. Acceler ated methods for nonconvex optimization. SIAM Journal on Optimization , 28(2):1751–1772, 2018

2018
[8]

Fercoq and Z

O. Fercoq and Z. Qu. Adaptive restart of accelerated grad ient methods under local quadratic growth condition. IMA Journal of Numerical Analysis , 39(4):2069–2095, 2019. 10

2019
[9]

Ghadimi and G

S. Ghadimi and G. Lan. Accelerated gradient methods for n onconvex nonlinear and stochastic programming. Mathematical Programming, 156:59–99, 2016

2016
[10]

Giselsson and S

P. Giselsson and S. Boyd. Monotonicity and restart in fa st gradient methods. In 53rd IEEE Conference on Decision and Control , pages 5058–5063. IEEE, 2014

2014
[11]

M. L. N. Gon¸ calves, J. G. Melo, and R. D. C. Monteiro. Imp roved pointwise iteration- complexity of a regularized ADMM and of a regularized non-Eu clidean hpe framework. SIAM Journal on Optimization , 27(1):379–407, 2017

2017
[12]

Guigues, J

V. Guigues, J. Liang, and R. D. C. Monteiro. Universal su bgradient and proximal bun- dle methods for convex and strongly convex hybrid composite optimization. arXiv preprint arXiv:2407.10073, 2024

2024
[13]

Krichene, A

W. Krichene, A. Bayen, and P. L. Bartlett. Accelerated m irror descent in continuous and discrete time. Advances in neural information processing systems , 28, 2015

2015
[14]

G. Lan, Z. Lu, and R. D. C. Monteiro. Primal-dual ﬁrst-or der methods with O(1/ǫ) iteration- complexity for cone programming. Mathematical Programming, 126(1):1–29, 2011

2011
[15]

Lemar´ echal

C. Lemar´ echal. An extension of davidon methods to non d iﬀerentiable problems. In Nondif- ferentiable optimization, pages 95–109. Springer, 1975

1975
[16]

Lemar´ echal

C. Lemar´ echal. Nonsmooth optimization and descent me thods. 1978

1978
[17]

Lessard, B

L. Lessard, B. Recht, and A. Packard. Analysis and desig n of optimization algorithms via integral quadratic constraints. SIAM Journal on Optimization , 26(1):57–95, 2016

2016
[18]

J. Liang. Primal-dual proximal bundle and conditional gradient methods for convex problems. arXiv preprint arXiv:2412.00585 , 2024

2024
[20]

Liang and R

J. Liang and R. D. C. Monteiro. An average curvature acce lerated composite gradient method for nonconvex smooth composite optimization problems. SIAM Journal on Optimization , 31(1):217–243, 2021

2021
[21]

Liang and R

J. Liang and R. D. C. Monteiro. A proximal bundle variant with optimal iteration-complexity for a large range of prox stepsizes. SIAM Journal on Optimization , 31(4):2955–2986, 2021

2021
[22]

Liang and R

J. Liang and R. D. C. Monteiro. Average curvature FISTA f or nonconvex smooth composite optimization problems. Computational Optimization and Applications , 86(1):275–302, 2023

2023
[24]

Liang, Renato D

J. Liang, Renato D. C. Monteiro, and C.-K. Sim. A FISTA-t ype accelerated gradient algorithm for solving smooth nonconvex composite optimization probl ems. Computational Optimization and Applications, 79(3):649–679, 2021

2021
[25]

H. Lin, J. Mairal, and Z. Harchaoui. A universal catalys t for ﬁrst-order optimization. Advances in Neural Information Processing Systems , 28, 2015. 11

2015
[26]

R. Miﬄin. A modiﬁcation and an extension of Lemar´ echal ’s algorithm for nonsmooth mini- mization. In Nondiﬀerential and variational techniques in optimizatio n, pages 77–90. Springer, 1982

1982
[27]

R. D. C. Monteiro, C. Ortiz, and B. F. Svaiter. An adaptiv e accelerated ﬁrst-order method for convex optimization. Computational Optimization and Applications , 64:31–73, 2016

2016
[29]

Necoara, Y

I. Necoara, Y. Nesterov, and F. Glineur. Linear converg ence of ﬁrst order methods for non- strongly convex optimization. Mathematical Programming, 175(1-2):69–107, 2019

2019
[30]

Nesterov

Y. Nesterov. A method for unconstrained convex minimiz ation problem with the rate of convergence O(1/k2). Doklady AN SSSR , 269:543–547, 1983

1983
[31]

Nesterov

Y. Nesterov. Introductory lectures on convex optimization : a basic cours e. Kluwer Academic Publ., Boston, 2004

2004
[32]

Nesterov

Y. Nesterov. Universal gradient methods for convex opt imization problems. Mathematical Programming, 152:381–404, 2015

2015
[33]

Nesterov

Y. Nesterov. Lectures on Convex Optimization , volume 137 of Springer Optimization and Its Applications. Springer, 2018

2018
[34]

O’Donoghue and E

B. O’Donoghue and E. J. Cand` es. Adaptive restart for ac celerated gradient schemes. Foun- dations of Computational Mathematics , 15(3):715–732, 2015

2015
[35]

W. Su, S. Boyd, and E. J. Cand` es. A diﬀerential equation f or modeling nesterov’s accelerated gradient method: Theory and insights. Journal of Machine Learning Research , 17(153):1–43, 2016

2016
[36]

Sujanani and R

A. Sujanani and R. D. C. Monteiro. Eﬃcient parameter-fr ee restarted accelerated gradient methods for convex and strongly convex optimization. arXiv preprint arXiv:2410.04248 , 2024

2024 arXiv
[37]

Wibisono, A

A. Wibisono, A. C. Wilson, and M. I. Jordan. A variationa l perspective on accelerated methods in optimization. proceedings of the National Academy of Sciences , 113(47):E7351–E7358, 2016

2016
[38]

P. Wolfe. A method of conjugate subgradients for minimi zing nondiﬀerentiable functions. In Nondiﬀerentiable optimization , pages 145–173. Springer, 1975. A Deferred proofs Proof of Lemma 2.1 : (a) This statement immediately follows from the recession s of τj and Aj in (3) and...

1975

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [28]

R. D. C. Monteiro and B. F. Svaiter. An accelerated hybri d proximal extragradient method for convex optimization and its implications to second-ord er methods. SIAM Journal on Optimization, 23(2):1092–1125, 2013

work page 2013

[2] [19]

A Doubly Accelerated Inexact Proximal Point Method for Nonconvex Composite Optimization Problems

J. Liang and R. D. C. Monteiro. A doubly accelerated inex act proximal point method for nonconvex composite optimization problems. Available on arXiv:1811.11378 , 2018

work page Pith review arXiv 2018

[3] [23]

Liang and R

J. Liang and R. D. C. Monteiro. A uniﬁed analysis of a clas s of proximal bundle methods for solving hybrid convex composite optimization problems. Mathematics of Operations Research, 49(2):832–855, 2024

work page 2024

[4] [1]

Ahn and S

K. Ahn and S. Sra. Understanding nesterov’s acceleratio n via proximal point method. In Sym- posium on Simplicity in Algorithms (SOSA) . Society for Industrial and Applied Mathematics, 2022

work page 2022

[5] [2]

Alamo, P

T. Alamo, P. Krupa, and D. Limon. Gradient based restart F ISTA. In 58th IEEE Conference on Decision and Control (CDC) , pages 3936–3941. IEEE, 2019

work page 2019

[6] [3]

Alamo, P

T. Alamo, P. Krupa, and D. Limon. Restart FISTA with globa l linear convergence. In 18th IEEE European Control Conference (ECC) , pages 1969–1974. IEEE, 2019

work page 1969

[7] [4]

Alamo, P Krupa, and D

T. Alamo, P Krupa, and D. Limon. Restart of accelerated ﬁr st order methods with linear convergence under a quadratic functional growth condition . IEEE Transactions on Automatic Control, 67(10):5200–5214, 2022

work page 2022

[8] [5]

Beck and M

A. Beck and M. Teboulle. A fast iterative shrinkage-thre sholding algorithm for linear inverse problems. SIAM Journal on Imaging Sciences , 2(1):183–202, 2009

work page 2009