betaFeedback

WorksComputer Science

SURFACE: Preference Distribution Matching on Pareto Fronts via Scalarization Steering

Abstract

Constructing a finite Pareto representation requires allocating a limited number of solutions across the Pareto front (PF). A desired preference distribution specifies this allocation by assigning more solutions to some PF regions and fewer to others. Scalarization generates Pareto solutions by varying objective weights and solving single-objective problems. However, a distribution of scalarization weights generally does not translate into the same distribution of Pareto solutions along the PF, because the weight-to-solution map is nonlinear and unknown in advance.To account for this distortion, we introduce SURFACE (Scalarization Updates for Redistributing Finite Approximations with Chart Estimation}), which adaptively updates scalarization weights to match a desired PF distribution. SURFACE estimates how each region of the weight grid corresponds to an area of the PF from the current scalarized solutions, then redistributes the weights accordingly. The method supports parallel, warm-started, inexact optimization with standard single-objective solvers. Theoretically, we establish that, under the stated regularity conditions, SURFACE achieves linear convergence of the preference distribution mismatch to a finite-resolution error floor of O(N−2/(M−1))\mathcal{O}(N^{-2/(M-1)}), where NN is the number of adaptive weights and MM is the number of objectives. Empirically, experiments across multiple benchmarks demonstrate improved target-mass balance and PF coverage compared with established baselines.

Introduction

Multi-objective optimization (MOO) is increasingly important in modern artificial intelligence (AI), including multimodal learning {(citep}(Wei & Hu, 2024; Saha et al., 2025)}, large language model (LLM) alignment {(citep}(Shi et al., 2024)}, and agentic AI systems {(citep}(Zhang et al., 2026)}. Let xx be a decision variable in the feasible domain X⊆Rn\mathcal{X} \subseteq\mathbb{R}^{n}. MOO minimizes a vector of objective functions fm:X→Rf_m : \mathcal{X} \to\mathbb{R}:

min⁡x∈Xf(x):=[f1(x),f2(x),…,fM(x)]T.(1)\min_{x \in\mathcal{X}} \mathbf{f}(x) := [f_1(x), f_2(x), \ldots, f_M(x)]^{\mathsf{T}} . \tag*{(1)}

In MOO, there is generally no single solution that minimizes all objectives simultaneously. Instead, the goal of (1) is to find a Pareto-optimal solution x⋆∈Xx^{\star} \in\mathcal{X}, for which no x∈Xx \in\mathcal{X} satisfies fm(x)≤fm(x⋆)f_m(x) \le f_m(x^{\star}) for all mm, with strict inequality for at least one mm. The collection of such solutions is called the Pareto set XPF\mathcal{X}_{\mathrm{PF}}. Its image FPF:=f(XPF)\mathcal{F}_{\mathrm{PF}} := f(\mathcal{X}_{\mathrm{PF}}) forms the Pareto front (PF) {(citep}(Miettinen, 1999)}.

Preference information is essential in MOO because Pareto optimality alone does not determine which tradeoffs matter most to a decision maker {(citep}(Pajasmaa et al., 2025)}. Under standard regularity conditions, the PF forms an (M−1)(M-1)-dimensional manifold of tradeoff solutions {(citep}(Hillermeier, 2001)}. This motivates two classical paradigms. A priori methods incorporate preferences before optimization to target preferred Pareto solutions {(citep}(Chen et al., 2024, 2025)}, but require preferences to be specified in advance. A posteriori methods instead construct a representative PF approximation for later selection {(citep}(Zhang et al., 2024; Boehmer & Wittmann, 2026)}, but must represent a continuous PF with finitely many solutions. This raises a fundamental question of how to allocate these solutions across the PF so that the resulting approximation reflects the desired preference distribution:

(Q1) How to generate Pareto solutions whose distribution matches a desired preference on the PF?

To specify the desired distribution of Pareto solutions, we prescribe a positive (unnormalized) preference profile ϑ:f(X)→(0,∞)\vartheta: f(\mathcal{X}) \to(0, \infty) over attainable objective values {(citep}(Wang et al., 2017; Chen et al., 2024). On the PF FPF⊆f(X)\mathcal{F}_{\mathrm{PF}} \subseteq f(\mathcal{X}), larger values of ϑ\vartheta indicate regions where more Pareto solutions are desired, with ϑ≡1\vartheta\equiv1 corresponding to uniform surface-area coverage. Existing methods seek Pareto stationarity, specified tradeoffs, or well-spread representations, but offer limited control over aligning a finite Pareto representation with ϑ\vartheta [22, 36, 48, 72]. To address this challenge, we build on scalarization, which generates Pareto solutions by solving single-objective problems parameterized by weights w\mathbf{w} in the (M−1)(M-1)-dimensional simplex ΔM\Delta_M [25, 40]. For example, weighted-sum scalarization yields x⋆(w)∈argminx∈X∑m=1Mwmfm(x)\mathbf{x}^{\star}(\mathbf{w}) \in\mathop{\mathrm{argmin}}_{\mathbf{x}\in\mathcal{X}} \sum_{m=1}^{M} w_m f_m(\mathbf{x}). Weighted Tchebycheff scalarization (TS) [58] can further recover Pareto solutions on nonconvex fronts [24, 38], as detailed in § 2. Under suitable regularity, varying w\mathbf{w} induces the weight-to-PF map [43, 24] (Hillermeier 2001):

w↦x⋆(w)↦fPF(w):=f(x⋆(w))∈FPF,w∈ΔM.(2)\mathbf{w} \mapsto\mathbf{x}^{\star}(\mathbf{w}) \mapsto\mathbf{f}_{\mathrm{PF}}(\mathbf{w}) := f(\mathbf{x}^{\star}(\mathbf{w})) \in\mathcal{F}_{\mathrm{PF}}, \quad\mathbf{w} \in\Delta_M . \tag*{(2)}

Although the weight simplex and the PF are both (M−1)(M-1)-dimensional, equally sized regions in weight space can correspond to regions of different surface areas on the PF [34]. We characterize this local area scaling under (2) by the factor J(w)J(\mathbf{w}). As illustrated in Figure 1, the change-of-variables formula combines J(w)J(\mathbf{w}) with the prescribed preference ϑ\vartheta to yield the target weight density:

Image not extracted

Figure 1. Preference ϑ(fPF(w))\vartheta(\mathbf{f}_{\mathrm{PF}}(\mathbf{w})) and the factor J(w)J(\mathbf{w}) measuring PF area scaling jointly determine the target weight density ρϑ(w)\rho_{\vartheta}(\mathbf{w}).

ρϑ(w)⏟target weight density∝ϑ(fPF(w))⏟PF preferencedet⁡(DfPF(w)⊤DfPF(w))⏟generalized PF traversal speed∝J(w).(3)\underbrace{\rho_{\vartheta}(\mathbf{w})}_{\text{target weight density}} \propto\underbrace{\vartheta(\mathbf{f}_{\mathrm{PF}}(\mathbf{w}))}_{\text{PF preference}} \underbrace{\sqrt{\det\left(D\mathbf{f}_{\mathrm{PF}}(\mathbf{w})^{\top}D\mathbf{f}_{\mathrm{PF}}(\mathbf{w})\right)}}_{\text{generalized PF traversal speed} \propto J(\mathbf{w})}. \tag*{(3)}

Here DfPF(w)D\mathbf{f}_{\mathrm{PF}}(\mathbf{w}) denotes the Jacobian of the weight-to-PF map (2). The square-root determinant generalizes the PF traversal speed in SURF [34]. § 2 provides the formal change-of-variables derivation and exact definitions of J(w)J(\mathbf{w}) and the normalized ρϑ(w)\rho_{\vartheta}(\mathbf{w}). Realizing this distribution with finitely many scalarization solves motivates our second question:

(Q2) How can we use finitely many scalarization weights to approximate the desired PF distribution under a limited computational budget?

We answer this question with SURFACE (Scalarization Updates for Redistributing Finite Approximations with Chart Estimation), a geometry-aware framework for matching finite Pareto representations to a desired preference distribution. As illustrated in Figure 2, SURFACE estimates the weight-to-PF geometry from a finite grid of scalarization weights and adaptively redistributes them toward the target weight density ρϑ\rho_{\vartheta} in (3).

Image not extracted

Figure 2. An overview of SURFACE. Cooler regions indicate lower target density, while warmer regions indicate higher density and a denser allocation of Pareto solutions. SURFACE estimates the weight-to-PF geometry and updates the scalarization weights to match this target distribution.

Specifically, it estimates the PF distribution induced by the weights by measuring the area of the PF patch corresponding to each cell of the weight grid. It then updates the weights so that the resulting Pareto solutions move toward the desired preference distribution.

In summary, our main contributions are:

C1) In § 2, we show how a desired preference distribution over the PF maps to a distribution of scalarization weights via the weight-to-PF geometry, and give sufficient conditions for Tchebycheff scalarization to yield a smooth, nondegenerate PF parameterization.

C2) In § 3, we introduce SURFACE, which adaptively updates scalarization weights to approximate a target preference distribution with a fixed number of Pareto solutions. It supports off-the-shelf optimizers, warm-started inexact solves, and parallel subproblem solves.

C3) In § 4, we provide a rigorous convergence analysis, showing that the weight updates contract the preference mismatch linearly down to a finite-resolution floor of O(N−2/(M−1))\mathcal{O}(N^{-2/(M-1)}).

C4) In Section 5, we empirically corroborate the effectiveness of SURFACE and verify our theoretical claims: across 9 baselines and benchmarks including ZDT/DTLZ and MO-Gymnasium, SURFACE consistently achieves strong, steady performance.

Related work. Existing approaches to finite PF representation mainly target well-spread solutions through different design criteria. Adaptive scalarization adjusts parameters toward more even objective-space spacing [26, 71, 52], but may require local sensitivity information [26] or sequential parameter updates [71, 52]. Geometric methods construct additional objective-space directions or constraints [18, 42, 27], while continuation methods trace the PF sequentially using local derivative or sensitivity information [48, 8] (Hillermeier 2001). Evolutionary methods adapt weights or reference points according to sparsity and diversity [69, 49, 19, 16], and finite-design methods use covering, packing, subset-selection, or repulsion objectives [6, 72, 55] (Borghi 2023). Preference-conditioned models instead learn mappings from preferences to Pareto solutions [44, 37, 65]. SURF [34] achieves uniform spacing on bi-objective PFs by inverting a cumulative distribution function (CDF) that measures the fraction of PF arc length reached as the weight increases, with density proportional to PF traversal speed. SURFACE extends this idea to higher-dimensional PFs and user-specified preferences by estimating PF geometry from parallel scalarized solves and adjusting weights accordingly. Appendix § 10.§ 1 provides detailed comparisons with existing methods.

Notation. Let ΔM:={w∈RM:wm≥0,∑m=1Mwm=1}\Delta_M := \{ \mathbf{w} \in\mathbb{R}^{M} : w_m \ge0, \sum_{m=1}^{M} w_m = 1 \} denote the weight simplex. We write DD for the Jacobian with respect to the displayed parameter coordinates, ∥⋅∥\lVert\cdot\rVert for the Euclidean norm, and ∥⋅∥∞\lVert\cdot\rVert_{\infty} for the uniform norm. For k≥0k \ge0, Ck,1C^{k,1} denotes functions with kk continuous derivatives and Lipschitz kkth derivative, with ∥⋅∥Ck\lVert\cdot\rVert_{C^k} the supremum norm through order kk. We use Vol⁡M−1\operatorname{Vol}_{M-1} for intrinsic weight-space volume. Recurring symbols are collected in Appendix § 10.§ 2.

Pareto-Front Geometry from Scalarization

Under suitable regularity, the weight-to-PF map (2) parameterizes the corresponding PF region [43, 24] (Hillermeier 2001). Sufficient conditions are established for TS in § 2.1 and weighted-sum in Appendix § 11.§ 4. This allows us to answer (Q1) by finding a distribution of scalarization weights whose induced Pareto outcomes fPF(w)f_{\mathrm{PF}}(\mathbf{w}) follow the target PF preference ϑ\vartheta. The key is to characterize how variations in w\mathbf{w} translate into changes on the PF.

To characterize how changes in w\mathbf{w} affect the PF, we use the surface-area change-of-variables formula. Following SURF [34 Appendix C.3], we define the PF-area factor that measures this local geometric scaling:

J(w):=1M⏟PF-areadet⁡(DPF(w)TDPF(w))⏟generalized PF traversal speed.(4)J(\mathbf{w}) := \underbrace{\frac{1}{\sqrt{M}}}_{\text{PF-area}} \underbrace{\sqrt{\det\left(D_{\mathrm{PF}}(\mathbf{w})^{\mathsf{T}}D_{\mathrm{PF}}(\mathbf{w})\right)}}_{\text{generalized PF traversal speed}} . \tag*{(4)}

Here DPFD_{\mathrm{PF}} is the Jacobian in M−1M-1 independent simplex coordinates, det⁡\det denotes intrinsic (M−1)(M-1)-dimensional simplex volume, and 1/M1/\sqrt{M} accounts for the corresponding volume conversion. A weight-space volume element dw\mathrm{d}\mathbf{w} thus corresponds to PF surface area J(w) dwJ(\mathbf{w})\,\mathrm{d}\mathbf{w}, so a simplex cell τ⊆ΔM\tau\subseteq\Delta_M represents area ∫τJ(w) dw\int_{\tau} J(\mathbf{w})\,\mathrm{d}\mathbf{w}, as illustrated in Figure 3. For M=2M=2, this recovers the PF traversal-speed measure used in SURF. To realize the target preference ϑ\vartheta, the probability assigned to each weight-space region τ\tau must be proportional to the preference-weighted PF area ∫τϑ(fPF(w))J(w) dw\int_{\tau}\vartheta(f_{\mathrm{PF}}(\mathbf{w}))J(\mathbf{w})\,\mathrm{d}\mathbf{w}. Normalization yields the target weight density

Weight-to-PF geometry

Figure 3. Weight-to-PF geometry. A simplex cell τ\tau maps to a PF region fPF(τ)f_{\mathrm{PF}}(\tau). The factor J(w)J(\mathbf{w}) generalizes PF traversal speed to local surface-area scaling, with ∫τJ(w) dw\int_{\tau} J(\mathbf{w})\,\mathrm{d}\mathbf{w} giving the area of the corresponding PF region.

ρϑ(w):=ϑ(fPF(w))J(w)∫ΔMϑ(fPF(v))J(v) dv.(5)\rho_{\vartheta}(\mathbf{w}) := \frac{\vartheta(f_{\mathrm{PF}}(\mathbf{w}))J(\mathbf{w})}{\int_{\Delta_M}\vartheta(f_{\mathrm{PF}}(\mathbf{v}))J(\mathbf{v})\,\mathrm{d}\mathbf{v}} . \tag*{(5)}

Thus, sampling weights from ρϑ\rho_{\vartheta} induces the prescribed PF distribution. In particular, ϑ≡1\vartheta\equiv1 yields ρ1∝J\rho_1 \propto J, recovering uniform PF surface-area coverage. To prepare for (Q2), since evaluating JJ may require costly sensitivity information, we use finitely many sampled weight–PF pairs {(wn,fPF(wn))}n=1N\{(\mathbf{w}_n,f_{\mathrm{PF}}(\mathbf{w}_n))\}_{n=1}^{N} to estimate the represented PF areas. The following lemma characterizes preference matching across their associated weight-space cells.

Lemma 1 (Preference Matching). Suppose fPFf_{\mathrm{PF}} is injective and has full tangent rank. Let C(W)\mathcal{C}(W) be a finite measurable partition of ΔM\Delta_M associated with the weights W={wn}n=1N⊆ΔMW=\{\mathbf{w}_n\}_{n=1}^{N}\subseteq\Delta_M, with cells of positive volume. For each τ∈C(W)\tau\in\mathcal{C}(W), let aτ:=fPF(τ)a_\tau:=f_{\mathrm{PF}}(\tau) denote the corresponding PF patch, and define R^ϑ(W):=max⁡τ∈C(W)∣∣C(W)∣∫ρϑ(w)dw−1)\widehat{R}_{\vartheta}(W):=\max_{\tau\in\mathcal{C}(W)}\left|\lvert\mathcal{C}(W)\rvert\int_{\rho_\vartheta(w)}\mathrm{d}w-1\right). Then R^ϑ(W)=0\widehat{R}_{\vartheta}(W)=0 if and only if all PF patches {aτ}τ∈C(W)\{a_\tau\}_{\tau\in\mathcal{C}(W)} carry equal preference-weighted area.

Appendix B.1 provides the proof of Lemma 1. The residual measures preference-mass imbalance across C(W)\mathcal{C}(W). The separate weight mesh used to reconstruct the PF is introduced in Section 3. For ϑ≡1\vartheta\equiv1, this residual measures PF surface-area imbalance.

Regular PF Charts from TS

In this section, we specialize the weight-to-PF map to TS and establish its PF coverage and regular chart properties. Appendix B.4 provides an alternative analysis for weighted-sum scalarization.

References

  1. [1]Axel Abels, Diederik Roijers, Tom Lenaerts, Ann Nowé, and Denis Steckelmacher. Dynamic weights in multi-objective deep reinforcement learning. In Proceedings of the 36th International Conference on Machine Learning, volume 97 of Proceedings of Machine Learning Research, pp. 11–20. PMLR, 2019.arxiv.org/abs/1809.07803
  2. [2]Lucas Nunes Alegre, Ana Bazzan, and Bruno C. da Silva. Optimistic linear support and successor features as a basis for optimal policy transfer. In Proceedings of the 39th International Conference on Machine Learning, volume 162 of Proceedings of Machine Learning Research, pp. 394–413. PMLR, 2022.arxiv.org/abs/2206.11326
  3. [3]Fernanda Beltrán, Oliver Cuate, and Oliver Schütze. The pareto tracer for general inequality constrained multi-objective optimization problems. Mathematical and Computational Applications, 25(4):80, 2020.DOI
  4. [4]Claude Berge. Topological spaces: Including a treatment of multi-valued functions, vector spaces and convexity. Oliver & Boyd, 1963.
  5. [5]Nicola Beume, Boris Naujoks, and Michael Emmerich. SMS-EMOA: Multiobjective selection based on dominated hypervolume. European Journal of Operational Research, 181(3):1653–1669, 2007.DOI
  6. [6]Julian Blank, Kalyanmoy Deb, Yashesh Dhebar, Sunith Bandaru, and Haitham Seada. Generating well-spaced points on a unit simplex for evolutionary many-objective optimization. IEEE Transactions on Evolutionary Computation, 25(1):48–60, 2020.DOI
  7. [7]Niclas Boehmer and Maximilian T Wittmann. Picking a representative set of solutions in multiobjective optimization: axioms, algorithms, and experiments. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 40, pp. 16717–16725, 2026.
  8. [8]Matthias Bolten, Onur Tanil Doganay, Hanno Gottschalk, and Kathrin Klamroth. Tracing locally pareto-optimal points by numerical integration. SIAM Journal on Control and Optimization, 59(5):3302–3328, 2021.arxiv.org/abs/2004.10820
  9. [9]J Frédéric Bonnans and Alexander Shapiro. Perturbation analysis of optimization problems. Springer Science & Business Media, 2013.
  10. [12]Martins Bruveris, Peter Michor, Adam Parusiński, and Armin Rainer. Moser’s theorem on manifolds with corners. Proceedings of the American Mathematical Society, 146(11):4889–4897, 2018.DOI
  11. [13]Chris J Budd and JF Williams. Moving mesh generation using the parabolic Monge–Ampère equation. SIAM Journal on Scientific Computing, 31(5):3438–3465, 2009.DOI
  12. [14]Lisha Chen, AFM Saif, Yanning Shen, and Tianyi Chen. FERERO: A flexible framework for preference-guided multi-objective learning. Advances in Neural Information Processing Systems, 37:18758–18805, 2024.arxiv.org/abs/2412.01773
  13. [15]Lisha Chen, Quan Xiao, Ellen Hidemі Fukuda, Xinyi Chen, Kun Yuan, and Tianyi Chen. Efficient first-order optimization on the pareto set for multi-objective learning under preference guidance. In Proceedings of the 42nd International Conference on Machine Learning, volume 267 of Proceedings of Machine Learning Research, pp. 9443–9486. PMLR, 2025.arxiv.org/abs/2504.02854
  14. [16]Ran Cheng, Yaochu Jin, Markus Olhofer, and Bernhard Sendhoff. A reference vector guided evolutionary algorithm for many-objective optimization. IEEE Transactions on Evolutionary Computation, 20(5):773–791, 2016.DOI
  15. [17]Bernard Dacorogna and Jürgen Moser. On a partial differential equation involving the jacobian determinant. Annales de l’Institut Henri Poincaré C, Analyse non linéaire, 7(1):1–26, 1990.DOI
  16. [18]Indraneel Das and John E Dennis. Normal-boundary intersection: A new method for generating the pareto surface in nonlinear multicriteria optimization problems. SIAM Journal on Optimization, 8(3):631–657, 1998.DOI
  17. [19]Kalyanmoy Deb and Himanshu Jain. An evolutionary many-objective optimization algorithm using reference-point-based nondominated sorting approach, part I: solving problems with box constraints. IEEE Transactions on Evolutionary Computation, 18(4):577–601, 2014.DOI
  18. [20]Kalyanmoy Deb, Amrit Pratap, Sameer Agarwal, and TAMT Meyarivan. A fast and elitist multi-objective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation, 6(2):182–197, 2002a.
  19. [21]Kalyanmoy Deb, Lothar Thiele, Marco Laumanns, and Eckart Zitzler. Scalable multi-objective optimization test problems. In Proceedings of the 2002 Congress on Evolutionary Computation (CEC), volume 1, pp. 825–830, Honolulu, Hawaii, USA, 2002b. IEEE.DOI
  20. [22]Jean-Antoine Désidéri. Multiple-gradient descent algorithm (MGDA) for multiobjective optimization. Comptes Rendus. Mathématique, 350(5-6):313–318, 2012.DOI
  21. [23]Qiang Du, Max D Gunzburger, and Lili Ju. Constrained centroidal voronoi tessellations for surfaces. SIAM Journal on Scientific Computing, 24(5):1488–1506, 2003.DOI
  22. [24]Matthias Ehrgott. Multicriteria optimization. Springer, 2005.
  23. [25]Matthias Ehrgott and Stefan Ruzika. Improved ε-constraint method for multiobjective programming. Journal of Optimization Theory and Applications, 138(3):375–396, 2008.DOI
  24. [26]Gabriele Eichfelder. An adaptive scalarization method in multiobjective optimization. SIAM Journal on Optimization, 19(4):1694–1718, 2009.DOI
  25. [27]Tohid Erfani and Sergei V Utyuzhnikov. Directed search domain: a method for even generation of the pareto frontier in multiobjective optimization. Engineering Optimization, 43(5):467–484, 2011.DOI
  26. [28]Ilyas Fatkhullin, Niao He, and Yifan Hu. Stochastic optimization under hidden convexity. SIAM Journal on Optimization, 35(4):2544–2571, 2025.arxiv.org/abs/2401.00108
  27. [29]Florian Felten, Lucas N Alegre, Ann Nowé, Ana Bazzan, El-Ghazali Talbi, Grégoire Danoy, and Bruno C. da Silva. A toolkit for reliable benchmarking and research in multi-objective reinforcement learning. Advances in Neural Information Processing Systems, 36:23671–23700, 2023.DOI
  28. [30]Anthony V Fiacco. Sensitivity analysis for nonlinear programming using penalty methods. Mathematical Programming, 10(1):287–311, 1976.DOI
  29. [32]Weizhang Huang and Robert D Russell. Adaptive moving mesh methods. Springer Science & Business Media, 2010.DOI
  30. [33]Hisao Ishibuchi, Hiroyuki Masuda, and Yusuke Nojima. A study on performance evaluation ability of a modified inverted generational distance indicator. In Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, pp. 695–702, 2015.DOI
  31. [34]Liuyuan Jiang, Chentong Huang, and Lisha Chen. SURF: Steering the scalarization weight to uniformly traverse the Pareto front. In Advances in Neural Information Processing Systems, 2026.arxiv.org/abs/2605.20619
  32. [35]Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In International Conference on Learning Representations, 2015.arxiv.org/abs/1412.6980
  33. [36]Xi Lin, Hui-Ling Zhen, Zhenhua Li, Qing-Fu Zhang, and Sam Kwong. Pareto multi-task learning. Advances in Neural Information Processing Systems, 32, 2019.arxiv.org/abs/1912.12854
  34. [37]Xi Lin, Zhiyuan Yang, Xiaoyuan Zhang, and Qingfu Zhang. Pareto set learning for expensive multi-objective optimization. Advances in Neural Information Processing Systems, 35:19231–19247, 2022.arxiv.org/abs/2210.08495
  35. [38]Xi Lin, Xiaoyuan Zhang, Zhiyuan Yang, Fei Liu, Zhenkun Wang, and Qingfu Zhang. Smooth tchebycheff scalarization for multi-objective optimization. In Proceedings of the 41st International Conference on Machine Learning, volume 235 of Proceedings of Machine Learning Research, pp. 30479–30509. PMLR, 2024.arxiv.org/abs/2402.19078
  36. [39]Yiping Liu, Hisao Ishibuchi, Naoki Masuyama, and Yusuke Nojima. Adapting reference vectors and scalarizing functions by growing neural gas to handle irregular pareto fronts. IEEE Transactions on Evolutionary Computation, 24(3):439–453, 2019.DOI
  37. [40]R Timothy Marler and Jasbir S Arora. The weighted sum method for multi-objective optimization: new insights. Structural and Multidisciplinary Optimization, 41(6):853–862, 2010.DOI
  38. [41]Benjamin Martin, Alexandre Goldsztejn, Laurent Granvilliers, and Christophe Jermann. On continuation methods for non-linear bi-objective optimization: towards a certified interval-based approach. Journal of Global Optimization, 64(1):3–16, 2016.
  39. [42]Achille Messac, Amir Ismail-Yahaya, and Christopher A Mattson. The normalized normal constraint method for generating the pareto frontier. Structural and Multidisciplinary Optimization, 25(2):86–98, 2003.DOI
  40. [43]Kaisa Miettinen. Nonlinear multiobjective optimization, volume 12. Kluwer Academic Publishers, Boston, MA, 1999.
  41. [44]Aviv Navon, Aviv Shamsian, Gal Chechik, and Ethan Fetaya. Learning the pareto front with hypernetworks. In International Conference on Learning Representations, 2021.arxiv.org/abs/2010.04104
  42. [45]Mahboubeb Nezhadmoghaddam, Julio Juárez, Jesús Guillermo Falcón-Cardona, Michael TM Emmerich, and André Deutz. Approximating optimal μ-point distributions over a generalized sphere for riesz s-energy using a gradient descent algorithm. Information Sciences, 747:123470, 2026.DOI
  43. [46]Juuso Pajasmaa, Kaisa Miettinen, and Johanna Silvennoinen. Group decision making in multiobjective optimization: A systematic literature review. Group Decision and Negotiation, 34(2):329–371, 2025.DOI
  44. [47]Victor Pereyra. Fast computation of equispaced pareto manifolds and pareto fronts for multiobjective optimization problems. Mathematics and Computers in Simulation, 79(6):1935–1947, 2009.
  45. [48]Victor Pereyra, Michael Saunders, and Jose Castillo. Equispaced pareto front construction for constrained bi-objective optimization. Mathematical and Computer Modelling, 57(9-10):2122–2131, 2013.
  46. [49]Yutao Qi, Xiaoliang Ma, Fang Liu, Licheng Jiao, Jianyong Sun, and Jianshe Wu. MOEA/D with adaptive weight adjustment. Evolutionary Computation, 22(2):231–264, 2014.DOI
  47. [50]Diederik M. Roijers, Peter Vamplew, Shimon Whiteson, and Richard Dazelely. A survey of multi-objective sequential decision-making. Journal of Artificial Intelligence Research, 48:67–113, 2013.
  48. [52]Bhupinder Singh Saini, Hemant Kumar Singh, Babooshka Shavazipour, and Kaisa Miettinen. An efficient iterative approach for uniformly representing pareto fronts. In International Conference on Evolutionary Multi-Criterion Optimization, pp. 241–256. Springer, 2025.DOI
  49. [53]John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347, 2017.arxiv.org/abs/1707.06347
  50. [54]Ke Shang, Hisao Ishibuchi, Linjun He, and Lie Meng Pang. A survey on the hypervolume indicator in evolutionary multiobjective optimization. IEEE Transactions on Evolutionary Computation, 25(1):1–20, 2020.DOI
  51. [55]Ke Shang, Tianye Shu, Hisao Ishibuchi, Yang Nan, and Lie Meng Pang. Benchmarking large-scale subset selection in evolutionary multi-objective optimization. Information Sciences, 622:755–770, 2023.DOI
  52. [56]Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Yang Wu, et al. DeepSeekMath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300, 2024.arxiv.org/abs/2402.03300
  53. [58]Ralph E Steuer and Eng-Ung Choo. An interactive weighted tchebycheff procedure for multiple objective programming. Mathematical Programming, 26(3):326–344, 1983.DOI
  54. [59]Ryoji Tanabe and Hisao Ishibuchi. An easy-to-use real-world multi-objective optimization problem suite. Applied Soft Computing, 89:106078, 2020.DOI
  55. [60]Ye Tian, Langchun Si, Xingyi Zhang, Kay Chen Tan, and Yaochu Jin. Local model-based pareto front estimation for multiobjective optimization. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 53(1):623–634, 2022.DOI
  56. [63]Holger Wendland. Scattered data approximation, volume 17. Cambridge university press Cambridge, 2005.DOI
  57. [64]Xiangmin Xu, Weizhang Huang, Robert D. Russell, and J. F. Williams. Convergence of de boor’s algorithm for the generation of equidistributing meshes. IMA Journal of Numerical Analysis, 31 (2):580–596, 2011.DOI
  58. [65]Rongguang Ye, Longcan Chen, Wei-Bin Kou, Jinyuan Zhang, and Hisao Ishibuchi. Pareto front shape-agnostic pareto set learning in multi-objective optimization. In 2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp. 4730–4736. IEEE, 2024a.arxiv.org/abs/2408.05778
  59. [66]Rongguang Ye, Longcan Chen, Jinyuan Zhang, and Hisao Ishibuchi. Evolutionary preference sampling for Pareto set learning. In Proceedings of the Genetic and Evolutionary Computation Conference, pp. 630–638, 2024b.DOI
  60. [67]Boxi Yu and Sergey Utyuzhnikov. Extension of the directed search domain algorithm for multi-objective optimization to higher dimensions. Annals of Operations Research, 340(2):1125–1160, 2024.DOI
  61. [69]Qingfu Zhang and Hui Li. MOEA/D: A multiobjective evolutionary algorithm based on decomposition. IEEE Transactions on Evolutionary Computation, 11(6):712–731, 2007.DOI
  62. [70]Sen Zhang, Jianwei Guo, Hui Zhang, Xiaohong Jia, Dong-Ming Yan, Junhai Yong, and Peter Wonka. Capacity constrained blue-noise sampling on surfaces. Computers & Graphics, 55:44–54, 2016.DOI
  63. [71]Wei Hong Zhang and Tong Gao. A min–max method with adaptive weightings for uniformly spaced pareto optimum points. Computers & Structures, 84(28):1760–1769, 2006.DOI
  64. [72]Xiaoyuan Zhang, Genghui Li, Xi Lin, Yichi Zhang, Yifan Chen, and Qingfu Zhang. Gliding over the pareto front with uniform designs. Advances in Neural Information Processing Systems, 37:2215–2245, 2024.DOI
  65. [73]Eckart Zitzler, Kalyanmoy Deb, and Lothar Thiele. Comparison of multiobjective evolutionary algorithms: Empirical results. Evolutionary Computation, 8(2):173–195, 2000.pubmed.ncbi.nlm.nih.gov/10843520

Paper details

vlnc:W_3fehs3sx
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 11

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Computer Science

generative AI tools

View statement
Looking for related papers…
Contents