Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

ConfigX: Modular Configuration for Evolutionary Algorithms via Multitask Reinforcement Learning

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read ConfigX claims that one pretrained policy can configure many evolutionary algorithms and zero-shot to unseen tasks.

desk verdict ConfigX's modular EA configuration is genuinely new, but the universal zero-shot claim falls because the reward and state need the global optimum f* (and true optimum x*), which real black-box problems don't provide. read the letter →

arxiv 2412.07507 v2 pith:3GYIGXOA submitted 2024-12-10 cs.LG cs.NE

Hongshu Guo , Zeyuan Ma , Jiacheng Chen , Yining Ma , Zhiguang Cao , Xinglin Zhang , Yue-Jiao Gong This is my paper · ORCID
classification cs.LGcs.NE
keywords meta-learningforblack-boxoptimizationevolutionaryalgorithmconfigurationmultitaskreinforcementlearningTransformerpolicymodulardesignzero-shotgeneralizationlifelongautomatic
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

ConfigX aims to answer whether a single pretrained agent can configure diverse evolutionary algorithms, rather than requiring one specialized controller per algorithm. The paper decomposes EAs into modular sub-modules, samples legal algorithm structures from the resulting space, and meta-learns a Transformer-based configuration policy through multitask reinforcement learning over a joint space of algorithms and problems. If the claim holds, automatic configuration becomes reusable: a user could adopt a new EA or problem without training a fresh policy. Experiments report zero-shot gains over a Bayesian-optimization configurator on both synthetic and real problem sets, plus faster fine-tuning on unseen algorithm families.

What carries the argument

The load-bearing machinery is the joint optimization task space $\mathcal{T} = \mathcal{A} \times \mathcal{I}$, where $\mathcal{A}$ is the space of legal algorithm structures generated by Modular-BBO, a hierarchical module system with controllable and uncontrollable sub-modules, topology rules, and a 16-bit module id, and $\mathcal{I}$ is the problem instance space. The policy is a Transformer that fuses each module id with nine landscape and status features, adds positional encodings, applies three blocks of module-aware multi-head self-attention, and decodes a per-module configuration. PPO optimizes the multitask objective $\frac{1}{KN}\sum_{i=1}^{KN}\sum_{t=1}^{H}\gamma^{t-1}R(s_t^i, c_t^i)$.

What would settle it

Run the exact ConfigX pipeline on a problem instance whose global optimum $f^*$ is withheld at runtime; the state features in Appendix C and the reward in Eq. (8) both normalize by $f^{0,*} - f^*$, so the state and reward are undefined unless $f^*$ is known. The claim of zero-shot generalization to unseen tasks would fail if the method cannot form a state or reward on any benchmark where $f^*$ is not supplied.

Watch

Extended reading notes

Core claim

ConfigX's central claim is that universal configuration is learnable: one policy, pretrained on a moderate set of optimization tasks, can dictate configurations for evolutionary algorithms it was never trained on and for problems outside its training distribution. The paper represents an EA as a sequence of typed sub-modules, each carrying a 16-bit module id and nine normalized optimization-status features, and lets a Transformer with module-aware self-attention read that sequence and emit per-module configuration distributions. Training maximizes the accumulated normalized improvement across many tasks with PPO. The authors report that the pretrained model zero-shots well within its algorithm family and to unseen real-world problems, outperforms the SMAC3 automatic configurator, and recovers strong performance on PSO/GA structures via fine-tuning when zero-shot transfer is weaker.

Load-bearing premise

ConfigX's reward and state features assume every problem's global optimum is known in advance, which is precisely what a black-box optimization problem hides.

Editorial extensions

If this is right

  • A user who wants to configure a new DE variant could plug it into Modular-BBO and use the pretrained model directly instead of running a fresh configuration search.
  • Pretraining on synthetic problems transfers to real problem suites such as protein docking and HPO-B, suggesting that industrial tasks can be configured without per-task tuning.
  • When a genuinely new algorithm family appears, the pretrained weights provide a warm start that reaches the same performance in roughly one-third to one-half the learning steps of retraining from scratch.
  • Dynamic per-step configuration can outperform a single static configuration found by Bayesian optimization, because the policy adapts parameters as optimization progresses.

Reading between the lines

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

  • Beyond the paper's tests, the reward and state features require the global optimum $f^*$; on a truly black-box problem with unknown $f^*$, ConfigX as specified cannot compute its own state or reward, so the method's demonstrated scope is benchmarks with known optima.
  • A natural extension not explored in the paper is training with a learned surrogate of the optimum, which would let the same architecture run in genuine black-box settings; that is the key experiment I would want next.
  • The module-aware attention suggests a diagnostic: checking which module tokens the policy attends to for DE versus PSO/GA could reveal whether weak zero-shot transfer comes from a genuinely shared representation or from family-specific shortcuts.
  • Because the algorithm space is hand-curated from existing EA literature, the universal claim is bounded by the module catalog; adding a genuinely new operator type still requires extending Modular-BBO, though fine-tuning may help.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes ConfigX, a MetaBBO framework that modularizes evolutionary algorithms into interchangeable sub-modules and trains a Transformer-based policy with multitask reinforcement learning to dynamically configure any algorithm structure sampled from this module space. The policy is pretrained on DE variants over BBOB instances and then evaluated zero-shot on in-distribution DE tasks, on out-of-distribution real-world problem benchmarks (protein docking and HPO-B), and on out-of-distribution PSO/GA algorithm structures. The paper also studies fine-tuning for lifelong learning and reports ablations of the Transformer components. The central claim is that a single pretrained configuration agent can generalize zero-shot to unseen tasks and outperform state-of-the-art baselines.

Significance. If the central claim were correct, ConfigX would be a substantive step toward a universal automatic configuration agent for evolutionary algorithms. The Modular-BBO system is a genuinely useful engineering contribution: the module catalog is large, the topology rules are practical, and the within-family zero-shot results on DE (Table 1, Figure 4, left) are empirically convincing. The paper also provides detailed appendices and additional experiments on training-set size and sub-module-set size. However, the central claim is undermined by an unacknowledged oracle dependency: the reward and the state features require the global optimum f* (and in two features the optimum location x*) of every problem instance. In genuine black-box optimization these quantities are unknown, so ConfigX cannot even form its observation vector on a truly unseen real problem, and the 'universal' and 'zero-shot' claims in the abstract are not supported. The comparison with SMAC3 is also not apples-to-apples because SMAC3 is used as a static configurator while ConfigX reconfigures at every step.

major comments (3)
  1. [§3.3, Eq. (8); Appendix C, Eqs. (9)–(16)] The reward defined in Eq. (8), rt = δ × (f*_{p,t−1} − f*_{p,t}) / (f*_{p,0} − f*_p), and the state features in Appendix C (Eqs. (9)–(11) and (15), which normalize objective values by f^{0,*} − f^*, and Eqs. (14) and (16), which require the distance to the optimum x*) all depend on the global optimum f* (and x*) of each problem instance. In black-box optimization f* and x* are unknown by definition, and the paper provides no estimation procedure. Consequently, ConfigX cannot compute its observation vector on any genuinely unseen problem, and the training reward is undefined outside benchmarks with known optima. This dependency is not disclosed as a limitation anywhere in the manuscript, and it directly undercuts the 'universal' and 'zero-shot generalization' claims in the abstract and in §3.2.
  2. [§4.2, observation 4; Figure 4, right] The paper's own results on T(2)test,out, the out-of-distribution algorithm structures (PSO/GA), show zero-shot performance that the authors describe as 'not as expected' and attribute to the structural difference between DE and PSO/GA. Yet the abstract claims 'robust zero-shot generalization to unseen tasks.' The evidence supports only within-family zero-shot transfer on DE, not cross-algorithm generalization. The claim should be restricted accordingly, and the discrepancy between the abstract and the experimental finding should be addressed.
  3. [Appendix E.2; §4.2] The SMAC3 baseline is used as a static configurator: it searches for a single configuration per algorithm for the entire optimization process, whereas ConfigX emits a new configuration at every optimization step. Because dynamic per-step configuration has strictly more expressive power, the comparison does not fairly establish that ConfigX 'outperforms state-of-the-art baselines.' The evaluation also omits any competing MetaBBO methods for dynamic algorithm configuration (e.g., GLEET or RL-DAS), which would be the most relevant comparators for the paper's stated contribution.
minor comments (6)
  1. [§4.1 vs. Appendix D, Algorithm 2] The experimental setup in §4.1 states that training lasts for 50 epochs, but Algorithm 2 in Appendix D iterates 'for epoch ← 1 to 100'. These numbers are inconsistent and should be aligned.
  2. [§3.3, Eq. (7); Table 3] The policy outputs a Gaussian distribution N(µ, Σ) over the configuration vector, but many configuration dimensions are discrete or categorical (e.g., operator selections in MULTISTRATEGY modules in Table 3). The paper does not specify how discrete choices are sampled or projected from the continuous Gaussian output, which leaves the action semantics underspecified.
  3. [Appendix C, Eqs. (9)–(11), (15)] The normalization expressions such as 'min{ f_i / f^{0,∗} − f^∗ }' should be written with parentheses, i.e., 'min{ f_i / (f^{0,∗} − f^∗) }', to avoid ambiguity about which quantity is in the denominator.
  4. [Figure 4] The y-axis of Figure 4 is not labeled; the reader must infer from Appendix E.2 that it shows the min-max normalized performance metric. Adding a y-axis label would improve readability.
  5. [§4.4] There is a typo in the ablation discussion: 'embdeddings' should be 'embeddings'.
  6. [Table 4] The text claims that SMAC3 'leads to a larger overall training time on all four task sets,' but Table 4 reports per-task-set training time, not a cumulative sum across task sets. The claim should be clarified or supported by the appropriate aggregate numbers.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ConfigX is an empirical RL system whose zero-shot claims are validated on held-out tasks; the f*-oracle dependence is a deployment limitation, not a circular step.

full rationale

ConfigX contains no mathematical derivation whose conclusion is assumed in its premises. The proposed pipeline (Modular-BBO sampling, transformer policy, PPO training) is evaluated against held-out task sets Ttest,in, T(1)_test,out, and T(2)_test,out, so the central 'zero-shot generalization' claim is an empirical outcome rather than an identity. The reward (Eq. 8) and several state features (Appendix C Eqs. 9-16) require the global optimum f* and optimal solution x*, which is a genuine and undisclosed limitation for true black-box deployment, but f* is an external benchmark oracle, not a fitted parameter or a renamed version of the predicted quantity; the same equations also define the training signal, not a post-hoc 'prediction' of the same fit. Self-citations (Guo et al. 2024; Ma et al. 2024a; Ma et al. 2024b; Ma et al. 2023) are used for the state-feature idea, the MetaBBO framing, and transformer-based configuration motifs, yet each component is restated in this paper and does not reduce the paper's own claims to an unverified citation. The paper's own admission that zero-shot transfer to PSO/GA is 'not as expected' (Section 4.2) further indicates the claim is falsifiable rather than forced. No circular step can be exhibited.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The framework does not introduce physical or mathematical entities. The main load-bearing assumptions are the sufficiency of the landscape features, the representativeness of the module space, and the availability of the global optimum. These are not validated outside the benchmark settings.

free parameters (2)
  • reward scale delta = 10
    In Eq. 8, delta=10 is chosen by hand to scale the reward and stabilize training across tasks. It affects the magnitude of policy gradients and is not derived from data.
  • maximum module count Lmax and maximum config size Cmax = not specified in text
    These padding constants are required to make the Transformer input and output shapes invariant across different algorithm structures, but their values are not given and could affect the model's capacity.
assumptions (5)
  • domain assumption The nine hand-designed landscape features (Appendix C) are sufficient to represent optimization status across different problem scopes and algorithm families.
    The features are borrowed from the authors' prior MetaBBO work and are used without a theoretical justification. The poor zero-shot transfer to PSO/GA suggests the representation is not sufficient for all EAs.
  • ad hoc to paper The topology rules in Modular-BBO generate a representative sample of legal EA structures, and the module space covers the important components of modern EAs.
    The specific set of modules and rules is curated by the authors; the paper provides no evidence that this space covers the diversity of EAs beyond the listed instances.
  • domain assumption PPO training with the task-agnostic reward of Eq. 8 converges to a policy that optimizes the true objective of Eq. 2.
    This is a standard RL assumption, but the reward requires the global optimum f* to be known, which is not available in genuine black-box settings.
  • domain assumption The global optimum f* is known exactly for every problem instance used for reward and state computation.
    This is required by Eq. 8 and Appendix C, but it is not stated as an assumption in the paper. For real-world black-box problems, f* is unknown, making the method inapplicable.
  • ad hoc to paper A Transformer with module-id tokens and positional encoding can represent the mapping from optimization state to effective configurations for any EA in the space.
    This is an architectural assumption. The failure on PSO/GA zero-shot indicates the model does not currently learn such a universal mapping.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ConfigX: Modular Configuration for Evolutionary Algorithms via Multitask Reinforcement Learning." pith.science (2026). https://pith.science/paper/3GYIGXOA

@misc{pith2026241207507,
  author       = {Pith},
  title        = {Pith review of: ConfigX: Modular Configuration for Evolutionary Algorithms via Multitask Reinforcement Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3GYIGXOA}},
  note         = {Machine review of arXiv:2412.07507}
}
read the original abstract

Recent advances in Meta-learning for Black-Box Optimization (MetaBBO) have shown the potential of using neural networks to dynamically configure evolutionary algorithms (EAs), enhancing their performance and adaptability across various BBO instances. However, they are often tailored to a specific EA, which limits their generalizability and necessitates retraining or redesigns for different EAs and optimization problems. To address this limitation, we introduce ConfigX, a new paradigm of the MetaBBO framework that is capable of learning a universal configuration agent (model) for boosting diverse EAs. To achieve so, our ConfigX first leverages a novel modularization system that enables the flexible combination of various optimization sub-modules to generate diverse EAs during training. Additionally, we propose a Transformer-based neural network to meta-learn a universal configuration policy through multitask reinforcement learning across a designed joint optimization task space. Extensive experiments verify that, our ConfigX, after large-scale pre-training, achieves robust zero-shot generalization to unseen tasks and outperforms state-of-the-art baselines. Moreover, ConfigX exhibits strong lifelong learning capabilities, allowing efficient adaptation to new tasks through fine-tuning. Our proposed ConfigX represents a significant step toward an automatic, all-purpose configuration agent for EAs.

Figures

Figures reproduced from arXiv: 2412.07507 by the authors.

Figure 1
Figure 1. Conceptual overview of different AC paradigms. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Left: The hierarchical polymorphism in Modular-BBO. Right: Legal/Illegal algorithm examples in Modular-BBO. Algorithm 1: Algorithm Structure Generation. Input: All accessible modules M, all Initialization modules Minit Output: A legal algorithm structure A. 1: Create an empty structure A = ∅, set index j = 0 2: Randomly select an Initialization module from Minit as aj 3: A ← A S aj 4: while not COMPLETED do 5: j = j… view at source ↗
Figure 3
Figure 3. The workflow of the Transformer based configuration policy in ConfigX. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Optimization curves of the pre-trained ConfigX model and the baselines, over three different zero-shot scenarios. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The learning curves of fine-tuning and re-training [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: The performance on advanced DE tasks. Ttrain Ttest,in T (1) test,out T (2) test,out ConfigX 527s / 7.03s – / 7.14s – / 8.64s – / 7.11s SMAC3 281s / 5.86s 286s / 5.93s – / 7.53s 284s / 5.89s Original – / 5.73s – / 5.76s – / 7.49s – / 5.72s Random – / 5.79s – / 5.80s – /…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. AutoSGNN: Automatic Propagation Mechanism Discovery for Spectral Graph Neural Networks

    cs.LG 2024-12 conditional novelty 5.0 of 10

    An LLM plus evolutionary-strategy framework that searches for per-dataset spectral GNN propagation mechanisms, with rank-1 sparse-split results but only rank-2 dense-split performance.

Reference graph

Works this paper leans on

89 extracted references · 74 canonical work pages · cited by 1 Pith paper

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Aleti, A.; and Moser, I. 2016. A systematic literature review of adaptive parameter control methods for evolutionary algorithms. ACM Comput. Surv

    2016

  4. [4]

    J.; Shamsi, M.; and Sedaaghi, M

    Amoshahy, M. J.; Shamsi, M.; and Sedaaghi, M. H. 2016. A novel flexible inertia weight particle swarm optimization algorithm. PloS one, 11(8): e0161558

    2016

  5. [5]

    Ans \'o tegui, C.; Sellmann, M.; and Tierney, K. 2009. A gender-based genetic algorithm for the automatic configuration of algorithms. In CP

    2009

  6. [6]

    P.; Jomaa, H

    Arango, S. P.; Jomaa, H. S.; Wistuba, M.; and Grabocka, J. 2021. HPO -B: A Large-Scale Reproducible Benchmark for Black-Box HPO based on Open ML . In NeurIPS

    2021

  7. [7]

    V.; Swiech, M.; Delgado, M.; and Neves-Jr, F

    Arruda, L. V.; Swiech, M.; Delgado, M.; and Neves-Jr, F. 2008. PID control of MIMO process based on rank niching genetic algorithm. Appl. Intell., 29(3): 290--305

    2008

  8. [8]

    L.; Kiros, J

    Ba, J. L.; Kiros, J. R.; and Hinton, G. E. 2016. Layer normalization. In NeurIPS

    2016

Show all 89 references
  1. [9]

    Baker, J. E. 2014. Adaptive selection methods for genetic algorithms. In Proceedings of the first international conference on genetic algorithms and their applications, 101--106. Psychology Press

    2014

  2. [10]

    D.; Das, S.; and Jalaian, B

    Biswas, S.; Saha, D.; De, S.; Cobb, A. D.; Das, S.; and Jalaian, B. A. 2021. Improving differential evolution through Bayesian hyperparameter optimization. In CEC, 832--840

    2021

  3. [11]

    S.; and Bo s kovi \'c , B

    Brest, J.; Mau c ec, M. S.; and Bo s kovi \'c , B. 2021. Self-adaptive differential evolution algorithm with population size reduction for single objective bound-constrained optimization: Algorithm j21. In CEC, 817--824. IEEE

    2021

  4. [12]

    Cao, Y.; and Shen, Y. 2020. Bayesian active learning for optimization and uncertainty quantification in protein docking. J. Chem. Theory Comput., 16(8): 5334--5347

    2020

  5. [13]

    Chen, A.; Dohan, D.; and So, D. 2024. EvoPrompting: language models for code-level neural architecture search. NeurIPS, 36

    2024

  6. [14]

    S.; and Suganthan, P

    Das, S.; Mullick, S. S.; and Suganthan, P. N. 2016. Recent advances in differential evolution--an updated survey. Swarm Evol. Comput., 27: 1--30

    2016

  7. [15]

    B.; et al

    Deb, K.; Agrawal, R. B.; et al. 1995. Simulated binary crossover for continuous search space. Complex systems, 9(2): 115--148

    1995

  8. [16]

    C.; Pearson, D

    Dobnikar, A.; Steele, N. C.; Pearson, D. W.; Albrecht, R. F.; Deb, K.; and Agrawal, S. 1999. A niched-penalty approach for constraint handling in genetic algorithms. In Artificial Neural Nets and Genetic Algorithms: Proceedings of the International Conference in Portoro z , Sl...

    1999

  9. [17]

    Eimer, T.; Biedenkapp, A.; Reimer, M.; Adriaensen, S.; Hutter, F.; and Lindauer, M. 2021. DACBench: A benchmark library for dynamic algorithm configuration. arXiv preprint arXiv:2105.08541

    2021 arXiv

  10. [18]

    Fialho, \'A . 2010. Adaptive operator selection for optimization. Ph.D. thesis, Universit \'e Paris Sud-Paris XI

    2010

  11. [19]

    Finn, C.; Abbeel, P.; and Levine, S. 2017. Model-agnostic meta-learning for fast adaptation of deep networks. In ICML

    2017

  12. [20]

    Gehring, J.; Auli, M.; Grangier, D.; Yarats, D.; and Dauphin, Y. N. 2017. Convolutional sequence to sequence learning. In ICML, 1243--1252. PMLR

    2017

  13. [21]

    E.; and Deb, K

    Goldberg, D. E.; and Deb, K. 1991. A comparative analysis of selection schemes used in genetic algorithms. In Foundations of genetic algorithms, volume 1, 69--93. Elsevier

    1991

  14. [22]

    Gong, W.; Fialho, A.; Cai, Z.; and Li, H. 2011. Adaptive strategy selection in differential evolution for numerical optimization: an empirical study. Inf. Sci., 181(24): 5364--5386

    2011

  15. [23]

    Guo, H.; Ma, Y.; Ma, Z.; Chen, J.; Zhang, X.; Cao, Z.; Zhang, J.; and Gong, Y.-J. 2024. Deep Reinforcement Learning for Dynamic Algorithm Selection: A Proof-of-Principle Study on Differential Evolution. TSMC

    2024

  16. [24]

    Gupta, A.; Fan, L.; Ganguli, S.; and Fei-Fei, L. 2022. Metamorph: learning universal controllers with transformers. In ICLR. ICLR

    2022

  17. [25]

    Halton, J. H. 1960. On the efficiency of certain quasi-random sequences of points in evaluating multi-dimensional integrals. Numerische Mathematik, 2: 84--90

    1960

  18. [26]

    Hansen, N.; Auger, A.; Finck, S.; and Ros, R. 2010. Real-parameter black-box optimization benchmarking 2010: Experimental setup. Ph.D. thesis, INRIA

    2010

  19. [27]

    Holland, J. H. 1992. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence. MIT press

    1992

  20. [28]

    Huang, C.; Li, Y.; and Yao, X. 2019. A survey of automatic parameter tuning methods for metaheuristics. TEC

    2019

  21. [29]

    Hwang, H.; Vreven, T.; Janin, J.; and Weng, Z. 2010. Protein--protein docking benchmark version 4.0. Proteins: Structure, Function, and Bioinformatics

    2010

  22. [30]

    M.; Das, S.; Ghosh, S.; Roy, S.; and Suganthan, P

    Islam, S. M.; Das, S.; Ghosh, S.; Roy, S.; and Suganthan, P. N. 2011. An adaptive differential evolution algorithm with novel mutation and crossover strategies for global numerical optimization. TSMC, 42(2): 482--500

    2011

  23. [31]

    Jansen, T. 2002. On the analysis of dynamic restart strategies for evolutionary algorithms. In PPSN, 33--43

    2002

  24. [32]

    Joe, S.; and Kuo, F. Y. 2008. Constructing Sobol sequences with better two-dimensional projections. SISC, 30(5): 2635--2654

    2008

  25. [33]

    Kadavy, T.; Viktorin, A.; Kazikova, A.; Pluhacek, M.; and Senkerik, R. 2023. Impact of boundary control methods on bound-constrained optimization benchmarking. In GECCO, 25--26

    2023

  26. [34]

    Kazimipour, B.; Li, X.; and Qin, A. K. 2014. A review of population initialization techniques for evolutionary algorithms. In CEC, 2585--2592. IEEE

    2014

  27. [35]

    Kennedy, J.; and Eberhart, R. 1995. Particle swarm optimization. In ICNN, volume 4, 1942--1948. IEEE

    1995

  28. [36]

    B.; Zhang, X.; Wang, H.; and Liu, J

    Li, X.; Wu, K.; Li, Y. B.; Zhang, X.; Wang, H.; and Liu, J. 2024. GLHF: General Learned Evolutionary Algorithm Via Hyper Functions. arXiv preprint arXiv:2405.03728

    2024 arXiv

  29. [37]

    Lian, H.; Ma, Z.; Guo, H.; Huang, T.; and Gong, Y.-J. 2024. RLEMMO: Evolutionary Multimodal Optimization Assisted By Deep Reinforcement Learning. In GECCO, 683--693

    2024

  30. [38]

    J.; Qin, A

    Liang, J. J.; Qin, A. K.; Suganthan, P. N.; and Baskar, S. 2006. Comprehensive learning particle swarm optimizer for global optimization of multimodal functions. TEC, 10(3): 281--295

    2006

  31. [39]

    Liang, J.-J.; and Suganthan, P. N. 2005. Dynamic multi-swarm particle swarm optimizer. In SIS, 124--129. IEEE

    2005

  32. [40]

    Lindauer, M.; Eggensperger, K.; Feurer, M.; Biedenkapp, A.; Deng, D.; Benjamins, C.; Ruhkopf, T.; Sass, R.; and Hutter, F. 2022. SMAC3: A Versatile Bayesian Optimization Package for Hyperparameter Optimization. JMLR, 23(54): 1--9

    2022

  33. [41]

    J.; and Sun, Y

    Liu, Q.; Du, S.; Van Wyk, B. J.; and Sun, Y. 2020. Niching particle swarm optimization based on Euclidean distance and hierarchical clustering for multimodal optimization. Nonlinear Dynamics, 99: 2459--2477

    2020

  34. [42]

    Lunacek, M.; and Whitley, D. 2006. The dispersion metric and the CMA evolution strategy. In GECCO, 477--484

    2006

  35. [43]

    Lynn, N.; and Suganthan, P. N. 2017. Ensemble particle swarm optimizer. Appl. Soft Comput., 55: 533--548

    2017

  36. [44]

    Ma, H.; Shen, S.; Yu, M.; Yang, Z.; Fei, M.; and Zhou, H. 2019. Multi-population techniques in nature inspired optimization algorithms: A comprehensive survey. Swarm Evol. Comput., 44: 365--387

    2019

  37. [45]

    Ma, Z.; Chen, J.; Guo, H.; and Gong, Y.-J. 2024 a . Neural exploratory landscape analysis. arXiv preprint arXiv:2408.10672

    2024 arXiv

  38. [46]

    Ma, Z.; Chen, J.; Guo, H.; Ma, Y.; and Gong, Y.-J. 2024 b . Auto-configuring Exploration-Exploitation Tradeoff in Evolutionary Computation via Deep Reinforcement Learning. In GECCO, 1497--1505

    2024

  39. [47]

    Ma, Z.; Guo, H.; Chen, J.; Li, Z.; Peng, G.; Gong, Y.-J.; Ma, Y.; and Cao, Z. 2023. MetaBox: A Benchmark Platform for Meta-Black-Box Optimization with Reinforcement Learning. In NeurIPS, volume 36

    2023

  40. [48]

    Ma, Z.; Guo, H.; Chen, J.; Peng, G.; Cao, Z.; Ma, Y.; and Gong, Y.-J. 2024 c . LLaMoCo: Instruction Tuning of Large Language Models for Optimization Code Generation. arXiv preprint arXiv:2403.01131

    2024 arXiv

  41. [49]

    Ma, Z.; Guo, H.; Gong, Y.-J.; Zhang, J.; and Tan, K. C. 2024 d . Toward Automated Algorithm Design: A Survey and Practical Guide to Meta-Black-Box-Optimization. arXiv preprint arXiv:2411.00625

    2024 arXiv

  42. [50]

    D.; Bashford, D.; Bellott, M.; Dunbrack Jr, R

    MacKerell Jr, A. D.; Bashford, D.; Bellott, M.; Dunbrack Jr, R. L.; Evanseck, J. D.; Field, M. J.; Fischer, S.; Gao, J.; Guo, H.; Ha, S.; et al. 1998. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J. Phys. Chem., 102(18): 3586--3616

    1998

  43. [51]

    Mahdavi, S.; Rahnamayan, S.; and Deb, K. 2016. Center-based initialization of cooperative co-evolutionary algorithm for large-scale optimization. In CEC, 3557--3565. IEEE

    2016

  44. [52]

    N.; Pan, Q.-K.; and Tasgetiren, M

    Mallipeddi, R.; Suganthan, P. N.; Pan, Q.-K.; and Tasgetiren, M. F. 2011. Differential evolution algorithm with ensemble of parameters and mutation strategies. Appl. Soft Comput., 11(2): 1679--1696

    2011

  45. [53]

    D.; Beckman, R

    McKay, M. D.; Beckman, R. J.; and Conover, W. J. 2000. A comparison of three methods for selecting values of input variables in the analysis of output from a computer code. Technometrics, 42(1): 55--61

    2000

  46. [54]

    Michalewicz, Z. 2013. Genetic algorithms+ data structures= evolution programs. Springer Science & Business Media

    2013

  47. [55]

    Peng, F.; Tang, K.; Chen, G.; and Yao, X. 2009. Multi-start JADE with knowledge transfer for numerical optimization. In CEC, 1889--1895. IEEE

    2009

  48. [56]

    Peng, H.; Han, Y.; Deng, C.; Wang, J.; and Wu, Z. 2021. Multi-strategy co-evolutionary differential evolution for mixed-variable optimization. KBS, 229: 107366

    2021

  49. [57]

    Peram, T.; Veeramachaneni, K.; and Mohan, C. K. 2003. Fitness-distance-ratio based particle swarm optimization. In SIS, 174--181. IEEE

    2003

  50. [58]

    G.; Wiehe, K.; Hwang, H.; Kim, B.-H.; Vreven, T.; and Weng, Z

    Pierce, B. G.; Wiehe, K.; Hwang, H.; Kim, B.-H.; Vreven, T.; and Weng, Z. 2014. ZDOCK server: interactive docking prediction of protein--protein complexes and symmetric multimers. Bioinformatics, 30(12): 1771--1773

    2014

  51. [59]

    Pol \'a kov \'a , R.; Tvrd \' k, J.; and Bujok, P. 2014. Controlled restart in differential evolution applied to CEC2014 benchmark functions. In CEC, 2230--2236. IEEE

    2014

  52. [60]

    E.; and Nielsen, R

    Pool, J. E.; and Nielsen, R. 2007. Population size changes reshape genomic patterns of diversity. Evolution, 61(12): 3001--3006

    2007

  53. [61]

    K.; and Suganthan, P

    Qin, A. K.; and Suganthan, P. N. 2005. Self-adaptive differential evolution algorithm for numerical optimization. In CEC, volume 2, 1785--1791. IEEE

    2005

  54. [62]

    Schulman, J.; Wolski, F.; Dhariwal, P.; Radford, A.; and Klimov, O. 2017. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347

    2017 arXiv

  55. [63]

    P.; and De Freitas, N

    Shahriari, B.; Swersky, K.; Wang, Z.; Adams, R. P.; and De Freitas, N. 2015. Taking the human out of the loop: A review of Bayesian optimization. Proceedings of the IEEE, 104(1): 148--175

    2015

  56. [64]

    M.; El-Saleh, A

    Shami, T. M.; El-Saleh, A. A.; Alswaitti, M.; Al-Tashi, Q.; Summakieh, M. A.; and Mirjalili, S. 2022. Particle swarm optimization: A comprehensive survey. IEEE Access, 10: 10031--10061

    2022

  57. [65]

    Sharma, M.; Komninos, A.; L \'o pez-Ib \'a \ n ez, M.; and Kazakov, D. 2019. Deep reinforcement learning based parameter control in differential evolution. In GECCO, 709--717

    2019

  58. [66]

    M.; and Mehrotra, D

    Shukla, A.; Pandey, H. M.; and Mehrotra, D. 2015. Comparative review of selection techniques in genetic algorithm. In ABLAZE, 515--519. IEEE

    2015

  59. [67]

    Sobol, I. 1967. The distribution of points in a cube and the accurate evaluation of integrals (in Russian) Zh. Vychisl. Mat. i Mater. Phys, 7: 784--802

    1967

  60. [68]

    Song, L.; Gao, C.; Xue, K.; Wu, C.; Li, D.; Hao, J.; Zhang, Z.; and Qian, C. 2024. Reinforced In-Context Black-Box Optimization. arXiv preprint arXiv:2402.17423

    2024 arXiv

  61. [69]

    Spears, W. M. 1995. Adapting Crossover in Evolutionary Algorithms. In Evolutionary Programming IV: Proceedings of the Fourth Annual Conference on Evolutionary Programming, 367

    1995

  62. [70]

    Stanovov, V.; Akhmedova, S.; and Semenkin, E. 2021. NL-SHADE-RSP algorithm with adaptive archive and selective pressure for CEC 2021 numerical optimization. In CEC, 809--816. IEEE

    2021

  63. [71]

    Storn, R.; and Price, K. 1997. Differential evolution--a simple and efficient heuristic for global optimization over continuous spaces. J. Glob. Optim., 11: 341--359

    1997

  64. [72]

    Sun, J.; Liu, X.; B \"a ck, T.; and Xu, Z. 2021. Learning adaptive differential evolution algorithm from optimization experiences by policy gradient. TEC, 25(4): 666--680

    2021

  65. [73]

    Tan, Z.; and Li, K. 2021. Differential evolution with mixed mutation strategy based on deep reinforcement learning. Appl. Soft Comput., 111: 107678

    2021

  66. [74]

    Tan, Z.; Tang, Y.; Li, K.; Huang, H.; and Luo, S. 2022. Differential evolution with hybrid parameters and mutation strategies based on reinforcement learning. Swarm Evol. Comput

    2022

  67. [75]

    Tanabe, R.; and Fukunaga, A. 2013. Success-history based parameter adaptation for differential evolution. In CEC, 71--78. IEEE

    2013

  68. [76]

    Tanabe, R.; and Fukunaga, A. S. 2014. Improving the search performance of SHADE using linear population size reduction. In CEC, 1658--1665. IEEE

    2014

  69. [77]

    Tao, X.; Li, X.; Chen, W.; Liang, T.; Li, Y.; Guo, J.; and Qi, L. 2021. Self-Adaptive two roles hybrid learning strategies-based particle swarm optimization. Inf. Sci., 578: 457--481

    2021

  70. [78]

    Tomassini, M.; Vanneschi, L.; Collard, P.; and Clergue, M. 2005. A study of fitness distance correlation as a difficulty measure in genetic programming. Evolutionary Computation, 13(2): 213--239

    2005

  71. [79]

    G.; and Gendreau, M

    Toulouse, M.; Crainic, T. G.; and Gendreau, M. 1996. Communication issues in designing cooperative multi-thread parallel searches. Springer

    1996

  72. [80]

    N.; Bischl, B.; and Torgo, L

    Vanschoren, J.; Van Rijn, J. N.; Bischl, B.; and Torgo, L. 2014. OpenML: networked science in machine learning. ACM SIGKDD Explorations Newsletter, 15(2): 49--60

    2014

  73. [81]

    N.; Kaiser, .; and Polosukhin, I

    Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. NeurIPS

    2017

  74. [82]

    Wang, Y.; Cai, Z.; and Zhang, Q. 2011. Differential evolution with composite trial vector generation strategies and control parameters. TEC, 15(1): 55--66

    2011

  75. [83]

    Williams, R. J. 1992. Simple statistical gradient-following algorithms for connectionist reinforcement learning. Machine Learning

    1992

  76. [84]

    Wu, D.; and Wang, G. G. 2022. Employing reinforcement learning to enhance particle swarm optimization methods. Engineering Optimization, 54(2): 329--348

    2022

  77. [85]

    Xue, K.; Xu, J.; Yuan, L.; Li, M.; Qian, C.; Zhang, Z.; and Yu, Y. 2022. Multi-agent dynamic algorithm configuration. NeurIPS, 35: 20147--20161

    2022

  78. [86]

    Yang, X.; Wang, R.; and Li, K. 2024. Meta-Black-Box Optimization for Evolutionary Algorithms: Review and Perspective. Available at SSRN 4956956

    2024

  79. [87]

    Ye, C.; Li, C.; Li, Y.; Sun, Y.; Yang, W.; Bai, M.; Zhu, X.; Hu, J.; Chi, T.; Zhu, H.; et al. 2023. Differential evolution with alternation between steady monopoly and transient competition of mutation strategies. Swarm Evol. Comput., 83: 101403

    2023

  80. [88]

    Zhabitskaya, E.; and Zhabitsky, M. 2013. Asynchronous differential evolution with restart. In NAA, 555--561. Springer

    2013

  81. [89]

    Zhang, J.; and Sanderson, A. C. 2009. JADE: adaptive differential evolution with optional external archive. TEC, 13(5): 945--958

    2009

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.