A Smoothing Consensus-Based Optimization Algorithm for Nonsmooth Nonconvex Optimization

Jiazhen Wei; Wei Bian

REVIEW 2 major objections 4 minor 39 references

A Smoothing Consensus-Based Optimization Algorithm for Nonsmooth Nonconvex Optimization

T0 review · 2 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a smoothing-based consensus optimization scheme converges almost surely to a common state for continuous nonconvex objectives and, under explicit parameter conditions, drives the objective value at that state…

desk verdict New finite-particle CBO convergence for nonsmooth objectives; consensus results are solid but the main error theorem has a genuinely flawed uniformity step. read the letter →

arxiv 2501.06804 v1 pith:2YLAMEDQ submitted 2025-01-12 math.OC

Jiazhen Wei , Wei Bian This is my paper · ORCID

classification math.OC MSC 90C2637N4065K05

keywords consensus-basedoptimizationsmoothingmethodnonsmoothnonconvexglobalconsensuserrorestimationnon-Lipschitzfiniteparticlesystemstochasticdifferentialequations

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

Consensus-based optimization (CBO) is a swarm method whose finite-particle convergence theory previously required smooth objectives or a mean-field limit. This paper proposes a smoothing variant, SCBO, that replaces the objective f with a smoothing function whose smoothing parameter decays to zero over time. It proves that, for any initial data, the particles converge almost surely to a common random state, and that under explicit drift, noise, and smoothing-schedule conditions the objective value at that state is bounded by fmin + E(β), with E(β) tending to zero as β tends to infinity. The result matters because it extends rigorous CBO convergence to continuous, nonconvex, possibly non-Lipschitz objective functions while working directly with the finite particle system.

What carries the argument

The load-bearing object is the smoothing function f̃(x, µ) of Definition 2.2: twice continuously differentiable in x, converging pointwise to f as µ decreases to 0, and satisfying the bounds |∇_µ f̃(x, µ)| ≤ κ $µ^{{-q}}$ and ‖∇²_xx f̃(x, µ)‖ ≤ η $µ^{{-q-1}}$ on bounded sets. In the algorithm, f̃ enters the weights exp(-β f̃(x_i(t), µ_t)) that define the consensus target x̄*(t), making those weights smooth enough for Ito's formula even though f itself is nonsmooth. The mechanism that carries the proof is the explicit separation formula for particle coordinates, which gives exponential decay; the Ito-formula inequality for the empirical average of the smoothed weights, which turns smoothing-parameter decay into a lower bound; and the Laplace principle, which converts that lower bound into the error estimate for f(x∞).

What would settle it

Take a continuous, non-Lipschitz objective such as f(x) = |x|^p on R with p in (0,1), choose initial data and parameters satisfying Assumptions 1-3 and inequality (45), and simulate the SCBO scheme; if the empirical essential infimum of f at the consensus state exceeds fmin + E(β), or if a positive fraction of paths leaves every bounded set, the central guarantee fails. A direct check of whether condition (39) holds for this example would settle whether the proof's uniformity assumption is valid.

Watch

Extended reading notes

Core claim

The central claim is stated as Theorem 4.2 and Corollary 4.1. For a continuous f mapping R^d to R_+ with a unique global minimizer, and for a smoothing function satisfying Definition 2.2, the SCBO dynamics produce a consensus state x∞ such that, whenever Assumptions 1-3 and inequality (45) hold, the essential infimum of f at x∞ is at most fmin + E(β), where E(β) tends to zero as β tends to infinity. Corollary 4.1 strengthens this to a guarantee: for any prescribed tolerance δ there exist parameters β, µ0, λ, and σ satisfying (45) with E(β) ≤ δ. The proof proceeds through an explicit log-difference formula that gives almost-sure and L² consensus, a martingale argument that yields the common limit x∞, and Ito's formula applied to the smoothed exponential weights, which produces a differential inequality that the Laplace principle converts into the final objective-error estimate.

Load-bearing premise

The error bound requires a single fixed bounded region that contains every particle path and the consensus target for all time, giving uniform smoothing constants; the paper proves only almost-sure convergence to a random limit, so the existence of that fixed region is not established.

Editorial extensions

If this is right

For any initial data, the SCBO particles reach a common consensus state almost surely whenever 2λ is greater than σ², with an explicit L² decay rate for pairwise disagreement.
For any target accuracy δ greater than zero, Corollary 4.1 guarantees parameters under which the objective value at the consensus state is within δ of the global minimum.
All sufficient conditions are dimension-independent, so the theoretical guarantees do not degrade as the search dimension grows.
An exponentially decaying smoothing schedule µ_t = µ_0 e^{-αt} with (q+1)α less than 2λ - σ² satisfies the key integrability assumption, giving an explicit tuning rule.

Reading between the lines

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

The proof's equation (39) assumes one fixed bounded set contains all particle paths and x̄*(t) almost surely for all time. Since the paper proves only almost-sure convergence to a random limit and does not assume bounded initial data, establishing a pathwise or high-probability bound on sup_t ‖x_i(t)‖ under moment conditions would make the expectation estimates unconditional.
Because the algorithm is gradient-free and the conditions are dimension-independent, the SCBO template is an attractive candidate for high-dimensional nonsmooth machine-learning objectives; the numerical experiments support good success rates, but the theory does not yet provide a complexity rate in dimension.
The same smoothing-by-convolution device could extend other stochastic swarm or metaheuristic algorithms that rely on Ito calculus to nonsmooth objectives, provided the smoothing parameter decays slowly enough relative to the consensus rate.

Share X Bluesky LinkedIn Reddit HN

Formalized claims in Lean

Claim #1: The central claim is stated as Theorem 4.2 and Corollary 4.1. For a continuous f mapping R^d to R_+ with a unique global minimizer, and for a smoothing function satisfying Definition 2.2, the SCBO dynamics produce a consensus state x∞ such that, whenever Assumptions 1-3 and inequality (45) hold, the essential infimum of f at x∞ is at most fmin + E(β), where E(β) tends to zero as β tends to infinit

/-- @claim 1 The central claim is stated as Theorem 4.2 and Corollary 4.1. For a continuous f mapping R^d to R_+ with a unique global minimizer, and for a smoothing function satisfying Definition 2.2, the SCBO dynamics produce a consensus state x∞ such that, whenever Assumptions 1-3 and inequality (45) hold, the essential infimum of f at x∞ is at most fmin + E(β), where E(β) tends to zero as β tends to infinit -/
def central_claim : Prop :=

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

Short version: the consensus half of this paper is a real step beyond Ha–Jin–Kim; the error theorem is not proved as stated, and the fix is doable.

What's new: they prove a.s. consensus and existence of a common consensus state for a CBO variant where f is merely continuous and possibly non-Lipschitz, using a smoothing function and a finite-particle SDE with common noise. The explicit formula for coordinate differences (geometric Brownian motion) does the heavy lifting, and Theorems 3.1 and 4.1 are clean and correct. That is genuinely new: the only prior finite-particle proof required C_b^2 and fixed beta. The idea of letting the smoothing parameter decay in time and tracking the weighted mean is natural, but the analysis is nontrivial.

The soft spot is Theorem 4.2, equation (39). They assert there is a bounded set X containing all paths and xbar*(t) a.s. for all t, so the smoothing constants kappa and eta are uniform. But Theorem 4.1 only gives a.s. convergence to a random limit; it does not give a deterministic bounded set. The stress-test note is right: even with bounded initial data, the geometric Brownian factors make P(sup_t |x_i_l(t) - xbar_l(t)| > R) > 0 for every R, so paths can leave any fixed ball on a set of positive probability. The set X would have to be random, and then kappa and eta are random and cannot be pulled out of the expectations in (40)-(43). So the error estimate does not follow as written. This is the load-bearing part of the paper's main claim, so it has to be fixed before the result is usable. The repair is probably adding bounded-support or strong moment assumptions on the initial data and proving a uniform-in-omega bound, or alternatively carrying kappa and eta inside the expectations and reworking condition (45). Corollary 4.1 is also over-stated: the proof chooses the initial data distribution as well as the parameters, so it is an existential statement over initial data too, and the text should say so.

Minor issue: the proof assumes finite second moments of the initial data (e.g., E[max_l (x_i_l(0)-bar_x_l(0))^2]) without stating it in Assumption 2.

Net: the consensus theorems are solid and worth keeping; the error analysis is conditional and needs revision. I would send this to a serious referee—the gap is identifiable and probably fixable, and the finite-particle nonsmooth CBO result is worth having in the literature.

For peer review: major revision, with a referee who knows SDE estimates.

Referee Report

2 major / 4 minor

Summary. The manuscript proposes SCBO, a finite-particle consensus-based optimization algorithm (Algorithm 1, Eqs. (14)-(15)) that replaces a nonsmooth, possibly non-Lipschitz objective f in (1) by a smoothing function tilde f(x, mu_t) with mu_t decaying to 0. It proves: (i) almost-sure global consensus for arbitrary initial data via an explicit geometric Brownian motion formula (23); (ii) existence of a common almost-sure limit x_infty (Theorem 4.1) under 2*lambda > sigma^2; (iii) an error estimate (Theorem 4.2) asserting ess inf f(x_infty) <= fmin + E(beta) with E(beta) -> 0 as beta -> infinity, provided condition (45) holds; and (iv) Corollary 4.1 claiming that for every delta > 0 parameters can be chosen to make E(beta) <= delta. Numerical experiments compare SCBO with a deterministic smoothing gradient method and with the CBO algorithm of [21].

Significance. If Theorems 4.2 and Corollary 4.1 were fully proved, the paper would make a meaningful contribution: it extends finite-particle CBO analysis to nonsmooth, non-Lipschitz objectives without passing to the mean-field limit, and it provides a quantitative error bound as beta -> infinity. The explicit GBM representation (23) is a clear strength, and the a.s. consensus theorem and the common-consensus-state theorem are supported by a direct, self-contained argument. The manuscript also gives reproducible numerical evidence that SCBO is competitive with existing methods. However, the central error estimate currently rests on an unproved uniform boundedness assertion, so the advertised error bound is not established as written.

major comments (2)

[Section 4.2, Eq. (39)] Equation (39) asserts that there exists a bounded set X such that xi(t) and xbar*(t) lie in X almost surely for all t. This does not follow from Theorem 4.1, which only proves almost-sure convergence of xi(t) to a random limit x_infty. Almost-sure convergence gives a sample-path-dependent eventual bound, not a deterministic bounded set. Moreover, for any nonzero initial difference, Eq. (23) shows that xi_l(t) - xj_l(t) is a geometric Brownian motion with negative drift; the supremum of such a process over [0, infinity) exceeds any fixed level R with positive probability. Hence no deterministic bounded set can contain the paths for almost all omega, even if the initial data are bounded. Because Definition 2.2(iii) provides kappa and eta only on a given bounded set, the constants kappa and eta used in (40)-(43) become omega-dependent if X is path-dependent, and they cannot be factored out of the expectations in (42)-(43) as written. Consequently, the derivation of (46)-(52), and hence Theorem 4.2 and Corollary 4.1, is not justified. A repair requires either a genuine uniform-in-omega bound on the paths (for example by adding a projection or truncation to Algorithm 1), or a reformulation of condition (45) using expectations of the random quantities kappa, eta, and tilde f_min.
[Section 4.1, proof of Theorem 4.1 and Eq. (33)] The proof of Theorem 4.1 uses (33) to conclude that E[integral_0^t (xi_l(s) - xbar*_l(s))^2 ds] is finite, and then applies the martingale convergence theorem to the stochastic integral. For (33) to be finite, one needs E[max_{1<=i<=N, 1<=l<=d} (xi_l(0) - xbar_l(0))^2] < infinity. Assumption 2 only says that the initial data are i.i.d. with common law xin and does not impose any moment condition. Thus the proof of the common-consensus-state theorem does not cover initial laws with infinite second moment, although the abstract and Theorem 4.1 claim convergence 'with any initial data'. Please either add a finite second-moment hypothesis to Assumption 2 and Theorem 4.1, or provide a localization argument that avoids the expectation bound.

minor comments (4)

[Section 3.2, proof of Theorem 3.1] The proof applies Ito's formula to ln x_ij_l(t) without treating the case x_ij_l(0)=0. If the initial difference is zero, the logarithm is undefined, but Eq. (23) remains valid by continuity (the difference is identically zero). Please handle this case separately before applying the log-transform.
[Section 4.2, Eq. (43)] In the estimate for E[Q3], the displayed intermediate inequality omits the factor 1/2 that appears in the definition of Q3 in Eq. (37). The final bound is still valid because dropping the factor 1/2 yields a weaker (more negative) lower bound, but the displayed inequality should be derived explicitly to avoid confusion.
[Corollary 4.1, proof] The phrase 'a suitable xin' in the proof is misleading: the initial law should be part of the problem data, not a parameter to be selected. In fact, since e^{beta(tilde f_min - f(xin))} <= 1 and converges to 1 as beta -> 0, a sufficiently small beta works for any fixed xin; please reword the proof so that the quantifier over the given initial law is clear.
[Abstract and Introduction] There are several typographical errors, including 'dose not' (should be 'does not') and broken spacing in 'focus es'. Please proofread the text.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the convergence and error estimates follow from a forward Lyapunov/Ito argument; condition (45) is a sufficient condition, not a restatement of the conclusion.

full rationale

The paper's claimed results are derived by explicit stochastic calculus rather than by importing the target as an assumption. Consensus follows from the closed-form difference formula (23); the common-limit theorem follows from martingale convergence after the exponential bound (32). The main error estimate in Theorem 4.2 is a forward lower bound on the Gibbs weight N^{-1} sum_i E[e^{-beta f~(x_i(t),mu_t)}] obtained from Ito's formula, Lemma 4.1, and separate estimates on Q1, Q2, Q3. Condition (45) is a genuine sufficient condition relating the initial Laplace transform E[e^{-beta f(x_in)}] to the smoothing parameters and the initial spread; it is not a restatement of the conclusion, and the final Laplace-principle step (Proposition 2.5) is an external result. The smoothing-function definition is cited from Bian-Chen but is restated and constructively justified in Remark 2.2, so the self-citation is not load-bearing. The only notable defect is the unproved assertion in (39) of a single deterministic bounded set containing all sample paths; almost-sure convergence to a random limit does not imply such a set, so kappa, eta, and f~_min may be path-dependent. That is a correctness gap in the proof of Theorem 4.2, not a circularity: the argument does not assume the conclusion, nor does any fitted value get renamed a prediction. Corollary 4.1's freedom to choose x_in near the minimizer weakens the force of the small-error statement, but the implication itself is not definitional.

Assumptions & free parameters 6 free parameters · 7 assumptions · 0 invented entities

The central claim rests on a smoothing-function existence assumption, a parameter and initial-data condition, and an unproven uniform boundedness premise. There are no new physical entities. The main hidden load is the deterministic bounded set X in equation (39), which is needed for the expectation estimates to be uniform.

free parameters (6)

beta
Inverse temperature in the exponential weights e^{-beta f_tilde}; user-chosen and central to the error bound E(beta). In Experiment 2 the authors set beta = 0.2.
mu_0
Initial smoothing parameter; user-chosen and required to be small to satisfy condition (56) in Corollary 4.1. Set to 0.0005 in Experiment 2.
epsilon
Slack parameter introduced in condition (45); hand-selected as a function of beta, mu_0, and delta to make the final error bound work.
lambda
Drift rate in the SDE; user-chosen, must satisfy 2*lambda > sigma^2. Corollary 4.1 chooses lambda and sigma together to control the term (2*lambda+sigma^2)*gamma.
sigma
Noise intensity in the SDE; user-chosen with the constraint 2*lambda > sigma^2. It affects the consensus rate and the error condition.
alpha
Decay rate of the smoothing parameter mu_t = mu_0 e^{-alpha t}; must satisfy Assumption 3. The numerical experiments use alpha = 0.1 or 0.9.

assumptions (7)

standard math Ito's formula and martingale convergence theorems.
Used throughout Sections 3 and 4 to compute dynamics for particle differences and for the smoothed exponential weights.
standard math Law of iterated logarithm for Brownian motion.
Used in Theorems 3.1 and 4.1 to show the linear decay term dominates Brownian fluctuations.
standard math Laplace principle as stated in Proposition 2.5.
Used to conclude E(beta) tends to 0 as beta tends to infinity; it requires f to be bounded and the initial measure to be absolutely continuous, which are not fully stated in Assumption 2.
domain assumption The objective f is continuous, nonnegative, and has a unique global minimizer.
These are the standing assumptions on problem (1), common in CBO analyses.
domain assumption There exists a smoothing function f_tilde satisfying Definition 2.2 with q in [0,1), including the Hessian bound involving mu^{-q-1}.
Needed for the Ito estimates. The paper shows such smoothing functions for Lipschitz f and special non-Lipschitz cases, but not for every continuous non-Lipschitz function.
ad hoc to paper There exists a deterministic bounded set X such that all particle paths and xbar* lie in X almost surely.
Equation (39) in Section 4.2 asserts this without proof; earlier theorems only give pathwise convergence to a random limit, so this uniform boundedness premise is not derived.
domain assumption Initial data have finite second moments and satisfy condition (45).
Assumption 2 alone does not imply finiteness of the maximum initial spread term, and condition (45) is the sufficient condition used in Theorem 4.2.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Smoothing Consensus-Based Optimization Algorithm for Nonsmooth Nonconvex Optimization." pith.science (2026). https://pith.science/paper/2YLAMEDQ

@misc{pith2026250106804,
  author       = {Pith},
  title        = {Pith review of: A Smoothing Consensus-Based Optimization Algorithm for Nonsmooth Nonconvex Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2YLAMEDQ}},
  note         = {Machine review of arXiv:2501.06804}
}

read the original abstract

Lately, a novel swarm intelligence model, namely the consensus-based optimization (CBO) algorithm, was introduced to deal with the global optimization problems. Limited by the conditions of Ito's formula, the convergence analysis of the previous CBO finite particle system mainly focuses on the problem with smooth objective function. With the help of smoothing method, this paper achieves a breakthrough by proposing an effective CBO algorithm for solving the global solution of a nonconvex, nonsmooth, and possible non-Lipschitz continuous minimization problem with theoretical analysis, which dose not rely on the mean-field limit. We indicate that the proposed algorithm exhibits a global consensus and converges to a common state with any initial data. Then, we give a more detailed error estimation on the objective function values along the state of the proposed algorithm towards the global minimum. Finally, some numerical examples are presented to illustrate the appreciable performance of the proposed method on solving the nonsmooth, nonconvex minimization problems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 35 canonical work pages

[21]

Ha, S.Y., Jin, S., Kim, D.: Convergence of a ﬁrst-order co nsensus-based global opti- mization algorithm. Math. Models Meth. Appl. Sci. 30, 2417–2444 (2020)

work page 2020
[1]

Springer, Berlin, Heridelberg (1985)

Bernt, O.: Stochastic Diﬀerential Equations: An Introdu ction with Applications. Springer, Berlin, Heridelberg (1985)

work page 1985
[2]

Bian, W., Chen, X.J.: W orst-case complexity of smoothing quadratic regularization methods for non-Lipschitzian optimization. SIAM J. Optim. 23(3), 1718–1741 (2013)

work page 2013
[3]

IEEE Trans

Bian, W., Chen, X.J.: Neural network for nonsmooth, nonco nvex constrained minimiza- tion via smooth approximation. IEEE Trans. Neural Netw. Lea rn. Syst. 25(3), 545–556 (2014)

work page 2014
[4]

Bian, W., Chen, X.J.: A smoothing proximal gradient algor ithm for nonsmooth convex regression with cardinality penalty. SIAM J. Numer. Anal. 58(1), 858–883 (2020)

work page 2020
[5]

Borghi, G., Herty, M., Pareschi, L.: Constrained consens us-based optimization. SIAM J. Optim. 33(1), 211–236 (2023)

work page 2023
[6]

Carrillo, J.A., Choi, Y.P., Totzeck, C., Tse, O.: An analy tical framework for a consensus- based global optimization method. Math. Models Meth. Appl. Sci. 28, 1037–1066 (2018)

work page 2018
[7]

ESAIM Cont rol Optim

Carrillo, J.A., Jin, S., Li, L., Zhu, Y.: A consensus-base d global optimization method for high dimensional machine learning problems. ESAIM Cont rol Optim. Calc. Var. 27(S5) (2021)

work page 2021

Show all 39 references

[8]

arXiv preprint arXiv: 2305.02894 (2023)

Carrillo, J.A., Trillos, N.G., Li, S., Zhu, Y.: FedCBO: Re aching group consensus in clus- tered federated learning through consensus-based optimiz ation. arXiv preprint arXiv: 2305.02894 (2023)

2023 arXiv
[9]

Chen, X.J.: Smoothing methods for nonsmooth, nonconvex m inimization. Math. Pro- gram. 134(1), 71–99 (2012)

2012
[10]

Chen, X.J., Kelley, C.T., Xu, F.M., Zhang, Z.K.: A smooth ing direct search method for Monte Carlo-based bound constrained composite nonsmoo th optimization. SIAM J. Sci. Comput. 40(4), A2174–A2199 (2018)

2018
[11]

Chen, X.J., Lu, Z.S., Pong, T.K.: Penalty methods for a cl ass of non-Lipschitz opti- mization problems. SIAM J. Optim. 26(3), 1465–1492 (2016) SCBO for nonsmooth nonconvex optimization 33

2016
[12]

arXiv preprint arXiv: 2311.08270 (2023)

Chenchene, E., Huang, H., Qiu, J.: A consensus-based alg orithm for non-convex multi- player games. arXiv preprint arXiv: 2311.08270 (2023)

2023 arXiv
[13]

Marcel- Dekker Inc., New York (1988)

Crow, E.L., Shimizu, K.: Lognormal Distributions: Theo ry and Application. Marcel- Dekker Inc., New York (1988)

1988
[14]

Springer, Berlin, Heidelberg (1998)

Dembo, A., Zeitouni, O.: Large Deviations Techniques an d Applications. Springer, Berlin, Heidelberg (1998)

1998
[15]

Demo, N., Tezzele, M., Rozza, G.: A supervised learning a pproach involving active subspaces for an eﬃcient genetic algorithm in high-dimensi onal optimization problems. SIAM J. Sci. Comput. 43(3), B831–B853 (2021)

2021
[16]

Fornasier, M., Huang, H., Pareschi, L., Sunnen, P.: Cons ensus-based optimization on the sphere II: Convergence to global minimizers and machine learning. J. Mach. Learn. Res. 22(237), 1–55 (2021)

2021
[17]

arXiv preprint arXiv: 2103.15130 (2022)

Fornasier, M., Klock, T., Riedl, K.: Consensus-based op timization methods converge globally. arXiv preprint arXiv: 2103.15130 (2022)

2022 arXiv
[18]

arXiv preprint arXiv: 2310.16610 (2024)

Fornasier, M., Richtarik, P., Ried, K., Sun, L.: Consens us-based optimization with trun- cated noise. arXiv preprint arXiv: 2310.16610 (2024)

2024 arXiv
[19]

Springer, Berlin (2013)

Gall, J.: Brownian Motion, Martingales, and Stochastic Calculus. Springer, Berlin (2013)

2013
[20]

IEEE Trans

Grefenstette, J.J.: Optimization of control parameter s for genetic algorithms. IEEE Trans. Syst. Man Cybern. -Syst. 16(1), 122–128 (1986)

1986
[22]

Ha, S.Y., Jin, S., Kim, D.: Convergence and error estimat es for time-discrete consensus- based optimization algorithms. Numer. Math. 147, 255–282 (2021)

2021
[23]

Hertlein, L., Ulbrich, M.: An inexact bundle algorithm f or nonconvex nonsmooth mini- mization in Hilbert space. SIAM J. Control Optim. 57(5), 3137–3165 (2019)

2019
[24]

Springer-Verlag, Berlin (1993)

Hiriart-Urruty, J.B., Lemarechal, C.: Convex Analysis and Minimization Algorithms I. Springer-Verlag, Berlin (1993)

1993
[25]

arXiv preprint arXiv: 2212.12334 (2023)

Huang, H., Qiu, J., Riedl, K.: Consensus-based optimiza tion for saddle point problems. arXiv preprint arXiv: 2212.12334 (2023)

2023 arXiv
[26]

Huang, J., Jiao, Y.L., Lu, X.L., Zhu, L.P.: Robust decodi ng from 1-bit compressive sampling with ordinary and regularized least squares. SIAM J. Sci. Comput. 40(4), A2062–A2086 (2018)

2018
[27]

Acta Appl

Hwang, C.R.: Simulated annealing: Theory and applicati ons. Acta Appl. Math. 12(1), 108–111 (1988)

1988
[28]

Springer, London (2008 )

Klenke, A.: Probability Theory. Springer, London (2008 )

2008
[29]

In: Proceedings of the 25th International Conference on Mac hine Learning

Meka, R., Jain, P., Caramanis, C., Dhillon, I.S.: Rank mi nimization via online learning. In: Proceedings of the 25th International Conference on Mac hine Learning. Helsinki, Finland (2008)

2008
[30]

Springer, NewYork (2014)

Pavliotis, G.A.: Stochastic Processes and Application s. Springer, NewYork (2014)

2014
[31]

Pinnau, R., Totzeck, C., Tse, O., Martin, S.: A consensus -based model for global op- timization and its mean-ﬁeld limit. Math. Models Meth. Appl . Sci. 27(1), 183–204 (2016)

2016
[32]

Spr inger, New York (1998)

Rockafellar, R., W ets, R.B.: Variational Analysis. Spr inger, New York (1998)

1998
[33]

John Wiley & Sons, New York (1995)

Ross, S.M.: Stochastic Processes. John Wiley & Sons, New York (1995)

1995
[34]

Ruszczy´ nski, A.: A stochastic subgradient method for n onsmooth nonconvex multilevel composition optimization. SIAM J. Control Optim. 59(3), 2301–2320 (2021)

2021
[35]

European J

Siarry, C.P.: Tabu search applied to global optimizatio n. European J. Oper. Res. 123(2), 256–270 (2000)

2000
[36]

Tarasewich, P., Mcmullen, P.R.: Swarm intelligence. Co mmun. ACM 45(8), 62–67 (2002)

2002
[37]

Totzeck, C., W olfram, M.T.: Consensus-based global opt imization with personal best. Math. Biosci. Eng. 17, 6026–6044 (2020)

2020
[38]

Soft Comput

Yang, X.S., Deb, S., Zhao, Y.X., Fong, S., He, X.S.: Swarm intelligence: Past, present and future. Soft Comput. 22, 5923–5933 (2018)

2018
[39]

Zhang, C., Chen, X.J.: Smoothing projected gradient met hod and its application to stochastic linear complementarity problems. SIAM J. Optim . 20(2), 627–649 (2009)

2009

Pith tools

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

[1] [21]

Ha, S.Y., Jin, S., Kim, D.: Convergence of a ﬁrst-order co nsensus-based global opti- mization algorithm. Math. Models Meth. Appl. Sci. 30, 2417–2444 (2020)

work page 2020

[2] [1]

Springer, Berlin, Heridelberg (1985)

Bernt, O.: Stochastic Diﬀerential Equations: An Introdu ction with Applications. Springer, Berlin, Heridelberg (1985)

work page 1985

[3] [2]

Bian, W., Chen, X.J.: W orst-case complexity of smoothing quadratic regularization methods for non-Lipschitzian optimization. SIAM J. Optim. 23(3), 1718–1741 (2013)

work page 2013

[4] [3]

IEEE Trans

Bian, W., Chen, X.J.: Neural network for nonsmooth, nonco nvex constrained minimiza- tion via smooth approximation. IEEE Trans. Neural Netw. Lea rn. Syst. 25(3), 545–556 (2014)

work page 2014

[5] [4]

Bian, W., Chen, X.J.: A smoothing proximal gradient algor ithm for nonsmooth convex regression with cardinality penalty. SIAM J. Numer. Anal. 58(1), 858–883 (2020)

work page 2020

[6] [5]

Borghi, G., Herty, M., Pareschi, L.: Constrained consens us-based optimization. SIAM J. Optim. 33(1), 211–236 (2023)

work page 2023

[7] [6]

Carrillo, J.A., Choi, Y.P., Totzeck, C., Tse, O.: An analy tical framework for a consensus- based global optimization method. Math. Models Meth. Appl. Sci. 28, 1037–1066 (2018)

work page 2018

[8] [7]

ESAIM Cont rol Optim

Carrillo, J.A., Jin, S., Li, L., Zhu, Y.: A consensus-base d global optimization method for high dimensional machine learning problems. ESAIM Cont rol Optim. Calc. Var. 27(S5) (2021)

work page 2021