The reviewed record of science sign in
Pith

arxiv: 2605.31016 · v1 · pith:XAN3AKME · submitted 2026-05-29 · cs.LG

An Efficient and Scalable Graph Condensation with Structure-Preserving

Yulin Hu , Fuyan Ou , Ye Yuan This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-06-29 00:01 UTCgrok-4.3pith:XAN3AKMErecord.jsonopen to challenge →

classification cs.LG
keywords graph condensationgraph neural networksstructure preservationdecoupled optimizationheat kernel propagationsynthetic graphsefficient graph methodsscalable condensation
0
0 comments X

The pith

SP-ESGC produces accurate synthetic graphs from large originals by decoupling node condensation from structure generation.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Graph condensation compresses big graphs into small synthetic ones so that GNNs can train and run with far less memory and time. Prior methods jointly optimize nodes and edges in one slow loop and often fail when the downstream GNN changes. SP-ESGC first diffuses node features with a heat kernel, clusters them into class centroids, then feeds the condensed nodes to a separate pre-trained edge predictor that copies structural patterns from the original graph. Experiments on real datasets show the resulting synthetic graphs match the accuracy of coupled baselines while running much faster and working across different GNN architectures.

Core claim

By separating node condensation (via heat-kernel propagation and hybrid clustering) from structure generation (via a pre-trained edge predictor), SP-ESGC achieves precise graph condensation at high computational efficiency and generalizes across diverse GNN architectures without requiring joint optimization.

What carries the argument

The decoupled design that separates node condensation from graph structure generation via a pre-trained edge predictor.

If this is right

  • Synthetic graphs can be generated in a single forward pass rather than iterative joint optimization.
  • The same condensed nodes can be paired with structures inferred by different predictors for different tasks.
  • Condensation time scales independently of the number of GNN architectures that will later use the output.
  • The method applies directly to large real-world graphs without prohibitive memory growth during condensation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Once the edge predictor is trained, condensation of new graphs in the same domain could reuse it without retraining.
  • The separation might let practitioners swap in stronger node-embedding methods while keeping the same structure predictor.
  • Similar decoupling could be tested on temporal or heterogeneous graphs if an appropriate pre-trained predictor is supplied.

Load-bearing premise

The pre-trained edge predictor can reliably infer transferable structural patterns from the original graph to produce accurate synthetic graphs without coupled optimization.

What would settle it

On a standard benchmark dataset, measure GNN test accuracy on SP-ESGC synthetic graphs versus coupled-optimization baselines; a large consistent drop in accuracy would falsify the claim of precise condensation.

Figures

Figures reproduced from arXiv: 2605.31016 by Fuyan Ou, Ye Yuan, Yulin Hu.

Figure 1
Figure 1. Figure 1: A schematic diagram of the proposed SP-ESGC framework. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The generalization capabilities of SP-ESGC on five datasets. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
read the original abstract

Graph condensation (GC) is pivotal for enabling Graph Neural Networks (GNNs) deployment in resource-constrained scenarios by compressing large-scale graphs into compact synthetic counterparts. Existing GC methods commonly suffer from computational inefficiency due to coupled optimization as well as encountering poor generalization across GNN architectures. To address these challenges, this study proposes an Efficient and Scalable Graph Condensation with Structure-Preserving (SP-ESGC), which possesses a decoupled design that separates node condensation from graph structure generation. Specifically, it first employs heat kernel feature propagation to generate node representation via spectral graph theory-inspired diffusion. Further, a novel hybrid clustering strategy is designed to extracts discriminative intra-class centroids from the node representation. Finally, a pre-trained edge predictor infers transferable structural patterns from the original graph, ensuring accurate synthetic graph generation. Extensive experiments on real-world graph datasets demonstrate that the proposed SP-ESGC implementes a precise GC with significantly high computational efficiency. Moreover, SP-ESGC also generalizes well across diverse GNN architectures.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The manuscript proposes SP-ESGC for graph condensation, using a decoupled pipeline: heat-kernel diffusion to generate node representations via spectral theory, hybrid clustering to extract intra-class centroids, and a separately pre-trained edge predictor to infer transferable structural patterns for the synthetic graph. The central claims are that this yields precise condensation (retaining GNN performance) with significantly higher computational efficiency than coupled-optimization baselines, plus strong generalization across GNN architectures, as shown by experiments on real-world datasets.

Significance. A working decoupled GC method could be significant for scaling GNNs to large graphs by avoiding the computational cost of joint node-structure optimization. The heat-kernel plus hybrid-clustering node step and the pre-trained predictor for structure are technically interesting if they deliver the claimed preservation of performance. However, the significance is currently limited because the provided text supplies no quantitative results, baselines, runtime numbers, or error analysis, and the key transferability assumption on the edge predictor lacks supporting derivation or bounds.

major comments (2)
  1. [Abstract] Abstract and method overview: the claims of 'precise GC with significantly high computational efficiency' and 'generalizes well across diverse GNN architectures' are asserted without any reported accuracy numbers, runtime comparisons, baseline methods, or statistical analysis, so the central empirical claims cannot be evaluated from the supplied text.
  2. [Method (edge predictor)] Method description of the edge-predictor stage: the pipeline decouples node condensation from edge generation and relies on a pre-trained predictor to reconstruct edges that preserve the properties needed for downstream GNN accuracy. No derivation, spectral bound, or topological guarantee is supplied showing that the predictor's output from the condensed node set will retain the necessary structure; the claim therefore rests entirely on an unverified empirical assertion of transferability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed feedback. We address the major comments point by point below, clarifying the role of the abstract and the empirical basis of the edge predictor while noting where the manuscript can be strengthened.

read point-by-point responses

  1. Referee: [Abstract] Abstract and method overview: the claims of 'precise GC with significantly high computational efficiency' and 'generalizes well across diverse GNN architectures' are asserted without any reported accuracy numbers, runtime comparisons, baseline methods, or statistical analysis, so the central empirical claims cannot be evaluated from the supplied text.

    Authors: The abstract serves as a high-level summary of the method and its intended benefits. The full manuscript includes a dedicated Experiments section reporting accuracy metrics, runtime comparisons against coupled baselines, results on multiple real-world datasets, and cross-GNN generalization tests with statistical details. If the version provided to the referee omitted these sections or tables, we will ensure the complete manuscript is supplied. We will revise the abstract to include one or two concrete performance highlights for better self-containment. revision: partial

  2. Referee: [Method (edge predictor)] Method description of the edge-predictor stage: the pipeline decouples node condensation from edge generation and relies on a pre-trained predictor to reconstruct edges that preserve the properties needed for downstream GNN accuracy. No derivation, spectral bound, or topological guarantee is supplied showing that the predictor's output from the condensed node set will retain the necessary structure; the claim therefore rests entirely on an unverified empirical assertion of transferability.

    Authors: The edge predictor is pre-trained on the original graph to capture transferable structural patterns and then applied to the condensed node set. The manuscript supports the approach through empirical results demonstrating that the resulting synthetic graphs preserve downstream GNN accuracy across architectures. We agree that no formal derivation, spectral bound, or topological guarantee is provided; the transferability claim is validated empirically rather than theoretically. This is a genuine limitation of the current work. We will add a paragraph in the method section discussing the empirical evidence, the decoupling assumptions, and the lack of theoretical bounds as an avenue for future analysis. revision: partial

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained

full rationale

The paper presents a decoupled three-stage pipeline (heat-kernel diffusion for node features, hybrid clustering for condensed nodes, and a separately pre-trained edge predictor for structure) whose outputs are not shown by any equation or self-citation to be definitionally equivalent to its inputs. No fitted parameter is relabeled as a prediction, no uniqueness theorem is imported from the authors' prior work, and the central efficiency and generalization claims rest on external empirical results rather than internal reductions. This is the normal non-circular case for a methods paper whose load-bearing step is an empirical assertion about the edge predictor's transferability.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Abstract-only review limits visibility into parameters or assumptions; no explicit free parameters or invented entities named.

axioms (2)
  • domain assumption Heat kernel feature propagation via spectral graph theory produces discriminative node representations suitable for clustering.
    Invoked as the first step of the method.
  • domain assumption A pre-trained edge predictor can extract transferable structural patterns independent of the specific GNN used downstream.
    Central to the decoupled structure generation step.

pith-pipeline@v0.9.1-grok · 5702 in / 1140 out tokens · 23238 ms · 2026-06-29T00:01:39.013879+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

85 extracted references · 11 canonical work pages · 3 internal anchors

  1. [1]

    Influence maximization in real-world closed social networks,

    S. Huang, W. Lin, Z. Bao, and J. Sun, “Influence maximization in real-world closed social networks,”arXiv preprint arXiv:2209.10286, 2022

    work page arXiv 2022

  2. [2]

    Do transformers really perform badly for graph representation?

    C. Ying, T. Cai, S. Luo, S. Zheng, G. Ke, D. He, Y . Shen, and T.-Y . Liu, “Do transformers really perform badly for graph representation?”Advances in neural information processing systems, vol. 34, pp. 28 877–28 888, 2021

    2021

  3. [3]

    Graph neural networks for social recommendation,

    W. Fan, Y . Ma, Q. Li, Y . He, E. Zhao, J. Tang, and D. Yin, “Graph neural networks for social recommendation,” inThe world wide web conference, 2019, pp. 417–426

    2019

  4. [4]

    Graph neural networks: Methods, applications, and opportunities,

    L. Waikhom and R. Patgiri, “Graph neural networks: Methods, applications, and opportunities,”arXiv preprint arXiv:2108.10733, 2021

    work page arXiv 2021

  5. [5]

    Nas-bench-graph: Benchmarking graph neural architecture search,

    Y . Qin, Z. Zhang, X. Wang, Z. Zhang, and W. Zhu, “Nas-bench-graph: Benchmarking graph neural architecture search,” Advances in neural information processing systems, vol. 35, pp. 54–69, 2022

    2022

  6. [6]

    Remember the past: Distilling datasets into addressable memories for neural networks,

    Z. Deng and O. Russakovsky, “Remember the past: Distilling datasets into addressable memories for neural networks,” Advances in Neural Information Processing Systems, vol. 35, pp. 34 391–34 404, 2022

    2022

  7. [7]

    Dataset Distillation

    T. Wang, J.-Y . Zhu, A. Torralba, and A. A. Efros, “Dataset distillation,”arXiv preprint arXiv:1811.10959, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  8. [8]

    Does graph distillation see like vision dataset counterpart?

    B. Yang, K. Wang, Q. Sun, C. Ji, X. Fu, H. Tang, Y . You, and J. Li, “Does graph distillation see like vision dataset counterpart?”Advances in Neural Information Processing Systems, vol. 36, pp. 53 201–53 226, 2023

    2023

  9. [9]

    Dataset condensation with differentiable siamese augmentation,

    B. Zhao and H. Bilen, “Dataset condensation with differentiable siamese augmentation,” inInternational Conference on Machine Learning. PMLR, 2021, pp. 12 674–12 685

    2021

  10. [10]

    arXiv preprint arXiv:2006.05929 (2020)

    B. Zhao, K. R. Mopuri, and H. Bilen, “Dataset condensation with gradient matching,”arXiv preprint arXiv:2006.05929, 2020

    work page arXiv 2006

  11. [11]

    Dataset condensation with distribution matching,

    B. Zhao and H. Bilen, “Dataset condensation with distribution matching,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2023, pp. 6514–6523

    2023

  12. [12]

    arXiv preprint arXiv:2011.00050 (2020)

    T. Nguyen, Z. Chen, and J. Lee, “Dataset meta-learning from kernel ridge-regression,”arXiv preprint arXiv:2011.00050, 2020

    work page arXiv 2011

  13. [13]

    Cafe: Learning to condense dataset by aligning features,

    K. Wang, B. Zhao, X. Peng, Z. Zhu, S. Yang, S. Wang, G. Huang, H. Bilen, X. Wang, and Y . You, “Cafe: Learning to condense dataset by aligning features,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 12 196–12 205

    2022

  14. [14]

    arXiv preprint arXiv:2110.07580 , year=

    W. Jin, L. Zhao, S. Zhang, Y . Liu, J. Tang, and N. Shah, “Graph condensation for graph neural networks,”arXiv preprint arXiv:2110.07580, 2021

    work page arXiv 2021

  15. [15]

    Condensing graphs via one-step gradient matching,

    W. Jin, X. Tang, H. Jiang, Z. Li, D. Zhang, J. Tang, and B. Yin, “Condensing graphs via one-step gradient matching,” in Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2022, pp. 720–730

    2022

  16. [16]

    arXiv preprint arXiv:2206.13697 , year=

    M. Liu, S. Li, X. Chen, and L. Song, “Graph condensation via receptive field distribution matching,”arXiv preprint arXiv:2206.13697, 2022

    work page arXiv 2022

  17. [17]

    Structure-free graph condensation: From large-scale graphs to condensed graph-free data,

    X. Zheng, M. Zhang, C. Chen, Q. V . H. Nguyen, X. Zhu, and S. Pan, “Structure-free graph condensation: From large-scale graphs to condensed graph-free data,”Advances in Neural Information Processing Systems, vol. 36, pp. 6026–6047, 2023

    2023

  18. [18]

    Navigating complexity: Toward lossless graph condensation via expanding window matching,

    Y . Zhang, T. Zhang, K. Wang, Z. Guo, Y . Liang, X. Bresson, W. Jin, and Y . You, “Navigating complexity: Toward lossless graph condensation via expanding window matching,”arXiv preprint arXiv:2402.05011, 2024

    work page arXiv 2024

  19. [19]

    Fast graph condensation with structure-based neural tangent kernel,

    L. Wang, W. Fan, J. Li, Y . Ma, and Q. Li, “Fast graph condensation with structure-based neural tangent kernel,” in Proceedings of the ACM Web Conference 2024, 2024, pp. 4439–4448

    2024

  20. [20]

    Kernel ridge regression-based graph dataset distillation,

    Z. Xu, Y . Chen, M. Pan, H. Chen, M. Das, H. Yang, and H. Tong, “Kernel ridge regression-based graph dataset distillation,” inProceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023, pp. 2850–2861

    2023

  21. [21]

    Simple graph condensation,

    Z. Xiao, Y . Wang, S. Liu, H. Wang, M. Song, and T. Zheng, “Simple graph condensation,” inJoint European Conference on Machine Learning and Knowledge Discovery in Databases. Springer, 2024, pp. 53–71

    2024

  22. [22]

    Rethinking and accelerating graph condensation: A training-free approach with class partition,

    X. Gao, G. Ye, T. Chen, W. Zhang, J. Yu, and H. Yin, “Rethinking and accelerating graph condensation: A training-free approach with class partition,” inProceedings of the ACM on Web Conference 2025, 2025, pp. 4359–4373

    2025

  23. [23]

    Graph spectral image smoothing using the heat kernel,

    F. Zhang and E. R. Hancock, “Graph spectral image smoothing using the heat kernel,”Pattern recognition, vol. 41, no. 11, pp. 3328–3342, 2008

    2008

  24. [24]

    Adaptive diffusion in graph neural networks,

    J. Zhao, Y . Dong, M. Ding, E. Kharlamov, and J. Tang, “Adaptive diffusion in graph neural networks,”Advances in neural information processing systems, vol. 34, pp. 23 321–23 333, 2021

    2021

  25. [25]

    Distributed implementation of heat kernel smoothing for graph signal denoising,

    C.-C. Tseng and S.-L. Lee, “Distributed implementation of heat kernel smoothing for graph signal denoising,” in2022 IEEE 65th International Midwest Symposium on Circuits and Systems (MWSCAS). IEEE, 2022, pp. 1–4

    2022

  26. [26]

    Fast approximate truncated svd,

    S. L. Shishkin, A. Shalaginov, and S. D. Bopardikar, “Fast approximate truncated svd,”Numerical Linear Algebra with Applications, vol. 26, no. 4, p. e2246, 2019

    2019

  27. [27]

    Random features for large-scale kernel machines,

    A. Rahimi and B. Recht, “Random features for large-scale kernel machines,”Advances in neural information processing systems, vol. 20, 2007

    2007

  28. [28]

    Disentangled condensation for large-scale graphs,

    Z. Xiao, Y . Wang, S. Liu, B. Hu, H. Wang, M. Song, and T. Zheng, “Disentangled condensation for large-scale graphs,” inProceedings of the ACM on Web Conference 2025, 2025, pp. 4494–4506

    2025

  29. [29]

    Herding dynamical weights to learn,

    M. Welling, “Herding dynamical weights to learn,” inProceedings of the 26th annual international conference on machine learning, 2009, pp. 1121–1128

    2009

  30. [30]

    Active Learning for Convolutional Neural Networks: A Core-Set Approach

    O. Sener and S. Savarese, “Active learning for convolutional neural networks: A core-set approach,”arXiv preprint arXiv:1708.00489, 2017

    work page internal anchor Pith review Pith/arXiv arXiv 2017

  31. [31]

    Semi-Supervised Classification with Graph Convolutional Networks

    T. Kipf, “Semi-supervised classification with graph convolutional networks,”arXiv preprint arXiv:1609.02907, 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  32. [32]

    Simplifying graph convolutional networks,

    F. Wu, A. Souza, T. Zhang, C. Fifty, T. Yu, and K. Weinberger, “Simplifying graph convolutional networks,” in International conference on machine learning. Pmlr, 2019, pp. 6861–6871

    2019

  33. [33]

    Graph attention networks,

    P. Velickovic, G. Cucurull, A. Casanova, A. Romero, P. Lio, Y . Bengioet al., “Graph attention networks,”stat, vol. 1050, no. 20, pp. 10–48 550, 2017

    2017

  34. [34]

    Inductive representation learning on large graphs,

    W. Hamilton, Z. Ying, and J. Leskovec, “Inductive representation learning on large graphs,”Advances in neural information processing systems, vol. 30, 2017

    2017

  35. [35]

    Predict then propagate: Graph neural networks meet personalized pagerank.arXiv preprint arXiv:1810.05997, 2018

    J. Gasteiger, A. Bojchevski, and S. G ¨unnemann, “Predict then propagate: Graph neural networks meet personalized pagerank,”arXiv preprint arXiv:1810.05997, 2018

    work page arXiv 2018

  36. [36]

    Graph tensor convolutional network,

    L. Wang, Y . Yuan, and X. Luo, “Graph tensor convolutional network,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2026

    2026

  37. [37]

    Advanced high-order graph convolutional networks with assorted time-frequency transforms,

    ——, “Advanced high-order graph convolutional networks with assorted time-frequency transforms,”IEEE/CAA Journal of Automatica Sinica, vol. 13, no. 2, pp. 394–408, 2026

    2026

  38. [38]

    A node-collaboration-informed graph convolutional network for highly accurate representation to undirected weighted graph,

    Y . Yuan, Y . Wang, and X. Luo, “A node-collaboration-informed graph convolutional network for highly accurate representation to undirected weighted graph,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 6, pp. 11 507–11 519, 2024

    2024

  39. [39]

    Gt-a 2 t: Graph tensor alliance attention network,

    L. Wang, K. Liu, and Y . Yuan, “Gt-a 2 t: Graph tensor alliance attention network,”IEEE/CAA Journal of Automatica Sinica, vol. 12, no. 10, pp. 2165–2167, 2024

    2024

  40. [40]

    A kalman-filter-incorporated latent factor analysis model for temporally dynamic sparse data,

    Y . Yuan, X. Luo, M. Shang, and Z. Wang, “A kalman-filter-incorporated latent factor analysis model for temporally dynamic sparse data,”IEEE Transactions on Cybernetics, vol. 53, no. 9, pp. 5788–5801, 2022

    2022

  41. [41]

    An adaptive divergence-based non-negative latent factor model,

    Y . Yuan, R. Wang, G. Yuan, and L. Xin, “An adaptive divergence-based non-negative latent factor model,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 53, no. 10, pp. 6475–6487, 2023

    2023

  42. [42]

    A multilayered-and-randomized latent factor model for high-dimensional and sparse matrices,

    Y . Yuan, Q. He, X. Luo, and M. Shang, “A multilayered-and-randomized latent factor model for high-dimensional and sparse matrices,”IEEE transactions on big data, vol. 8, no. 3, pp. 784–794, 2020

    2020

  43. [43]

    A generalized and fast-converging non-negative latent factor model for predicting user preferences in recommender systems,

    Y . Yuan, X. Luo, M. Shang, and D. Wu, “A generalized and fast-converging non-negative latent factor model for predicting user preferences in recommender systems,” inProceedings of The Web Conference 2020, 2020, pp. 498–507

    2020

  44. [44]

    Temporal web service qos prediction via kalman filter-incorporated latent factor analysis

    Y . Yuan, M. Shang, and X. Luo, “Temporal web service qos prediction via kalman filter-incorporated latent factor analysis.” inECAI, 2020, pp. 561–568

    2020

  45. [45]

    Non-gradient hash factor learning for high-dimensional and incomplete data representation learning,

    D. Wu, S. Li, Y . He, X. Luo, and X. Gao, “Non-gradient hash factor learning for high-dimensional and incomplete data representation learning,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    2026

  46. [46]

    Learning accurate representation to nonstandard tensors via a mode-aware tucker network,

    H. Wu, Q. Wang, X. Luo, and Z. Wang, “Learning accurate representation to nonstandard tensors via a mode-aware tucker network,”IEEE Transactions on Knowledge and Data Engineering, 2025

    2025

  47. [47]

    Tensor low-rank orthogonal compression for convolutional neural networks,

    Y . He and X. Luo, “Tensor low-rank orthogonal compression for convolutional neural networks,”IEEE/CAA Journal of Automatica Sinica, vol. 13, no. 1, pp. 227–229, 2026

    2026

  48. [48]

    Discovering spatiotemporal–individual coupled features from nonstandard tensors—a novel dynamic graph mixer approach,

    F. Bi, T. He, Y .-S. Ong, and X. Luo, “Discovering spatiotemporal–individual coupled features from nonstandard tensors—a novel dynamic graph mixer approach,”IEEE Transactions on Neural Networks and Learning Systems, 2025

    2025

  49. [49]

    Scg: A novel spatiotemporal coupling graph convolutional network-incorporated approach for dynamic qos estimation,

    F. Bi and T. He, “Scg: A novel spatiotemporal coupling graph convolutional network-incorporated approach for dynamic qos estimation,” in2024 IEEE International Conference on Systems, Man, and Cybernetics (SMC). IEEE, 2024, pp. 635–640

    2024

  50. [50]

    A convolution bias-incorporated nonnegative latent factorization of tensors model for accurate representation learning to dynamic directed graphs,

    Q. Wang, H. Wu, and X. Luo, “A convolution bias-incorporated nonnegative latent factorization of tensors model for accurate representation learning to dynamic directed graphs,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2025

    2025

  51. [51]

    Knowledge-driven multiple instance learning with hierarchical cluster-incorporated aware filtering for larynx pathological grading,

    C. Li, P. Huang, J. Qin, and X. Luo, “Knowledge-driven multiple instance learning with hierarchical cluster-incorporated aware filtering for larynx pathological grading,”IEEE Journal of Biomedical and Health Informatics, 2025

    2025

  52. [52]

    Fmvpci: a multiview fusion neural network for identifying protein complex via fuzzy clustering,

    Y . Yang, L. Hu, G. Li, D. Li, P. Hu, and X. Luo, “Fmvpci: a multiview fusion neural network for identifying protein complex via fuzzy clustering,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2025

    2025

  53. [53]

    Graph-based prediction of mirna-drug associations with multisource information and metapath enhancement matrices,

    M.-Y . Wu, P.-W. Hu, Z.-H. You, J. Zhang, L. Hu, and X. Luo, “Graph-based prediction of mirna-drug associations with multisource information and metapath enhancement matrices,”IEEE Journal of Biomedical and Health Informatics, 2025

    2025

  54. [54]

    Fuzzy mixture-of-experts aggregation for organoid identification with multiscale state space features,

    X. Deng, P. Hu, T. Herget, F. Tan, X. Zhu, J. Zhang, Y .-a. Huang, L. Hu, Z. You, and X. Luo, “Fuzzy mixture-of-experts aggregation for organoid identification with multiscale state space features,”IEEE Transactions on Fuzzy Systems, vol. 34, no. 1, pp. 324–335, 2025

    2025

  55. [55]

    Ncsac: Effective neural community search via attribute-augmented conductance,

    L. Lin, Q. Li, M. Qiao, Z. Wang, J. Zhao, R.-H. Li, X. Luo, and T. Jia, “Ncsac: Effective neural community search via attribute-augmented conductance,”IEEE Transactions on Knowledge and Data Engineering, vol. 38, no. 2, pp. 1221–1235, 2025

    2025

  56. [56]

    A proximal-admm-incorporated nonnegative latent-factorization-of-tensors model for representing dynamic cryptocurrency transaction network,

    X. Liao, H. Wu, T. He, and X. Luo, “A proximal-admm-incorporated nonnegative latent-factorization-of-tensors model for representing dynamic cryptocurrency transaction network,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2025

    2025

  57. [57]

    Multi-scale collaborative distillation graph neural networks for session-based recommendation,

    J. Gou, Y . Cheng, B. Ma, L. Du, X. Luo, and Z. Yi, “Multi-scale collaborative distillation graph neural networks for session-based recommendation,”IEEE Transactions on Services Computing, 2025

    2025

  58. [58]

    A generalized nesterov’s accelerated gradient-incorporated non-negative latent- factorization-of-tensors model for efficient representation to dynamic qos data,

    M. Chen, R. Wang, Y . Qiao, and X. Luo, “A generalized nesterov’s accelerated gradient-incorporated non-negative latent- factorization-of-tensors model for efficient representation to dynamic qos data,”IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 8, no. 3, pp. 2386–2400, 2024

    2024

  59. [59]

    Symmetry and graph bi-regularized non-negative matrix factorization for precise community detection,

    Z. Liu, X. Luo, and M. Zhou, “Symmetry and graph bi-regularized non-negative matrix factorization for precise community detection,”IEEE Transactions on Automation Science and Engineering, vol. 21, no. 2, pp. 1406–1420, 2023

    2023

  60. [60]

    Mmlf: Multi-metric latent feature analysis for high-dimensional and incomplete data,

    D. Wu, P. Zhang, Y . He, and X. Luo, “Mmlf: Multi-metric latent feature analysis for high-dimensional and incomplete data,”IEEE transactions on services computing, vol. 17, no. 2, pp. 575–588, 2023

    2023

  61. [61]

    Learning error refinement in stochastic gradient descent-based latent factor analysis via diversified pid controllers,

    J. Li, Y . Yuan, and X. Luo, “Learning error refinement in stochastic gradient descent-based latent factor analysis via diversified pid controllers,”IEEE Transactions on Emerging Topics in Computational Intelligence, 2025

    2025

  62. [62]

    A fuzzy pid-incorporated stochastic gradient descent algorithm for fast and accurate latent factor analysis,

    Y . Yuan, J. Li, and X. Luo, “A fuzzy pid-incorporated stochastic gradient descent algorithm for fast and accurate latent factor analysis,”IEEE Transactions on Fuzzy Systems, vol. 32, no. 7, pp. 4049–4061, 2024

    2024

  63. [63]

    Adaptively-accelerated parallel stochastic gradient descent for high-dimensional and incomplete data representation learning,

    W. Qin, X. Luo, and M. Zhou, “Adaptively-accelerated parallel stochastic gradient descent for high-dimensional and incomplete data representation learning,”IEEE Transactions on Big Data, vol. 10, no. 1, pp. 92–107, 2023

    2023

  64. [64]

    Sgd-dyg: Self-reliant global dependency apprehending on dynamic graphs,

    M. Han, L. Wang, Y . Yuan, and X. Luo, “Sgd-dyg: Self-reliant global dependency apprehending on dynamic graphs,” in Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V . 2, 2025, pp. 802–813

    2025

  65. [65]

    Auto-encoding neural tucker factorization,

    P. Tang, X. Luo, and J. Woodcock, “Auto-encoding neural tucker factorization,”IEEE Transactions on Knowledge and Data Engineering, 2025

    2025

  66. [66]

    Multimetric autoencoder for representing high-dimensional and incomplete data,

    D. Wu, C. Liang, Y . He, Y . Qiao, and X. Luo, “Multimetric autoencoder for representing high-dimensional and incomplete data,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 56, no. 3, pp. 1533–1546, 2026

    2026

  67. [67]

    A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data,

    X. Liao, H. Wu, and X. Luo, “A novel tensor causal convolution network model for highly-accurate representation to spatio-temporal data,”IEEE Transactions on Automation Science and Engineering, 2025

    2025

  68. [68]

    Adaptive divergence-based non-negative latent factor analysis of high-dimensional and incomplete matrices from industrial applications,

    Y . Yuan, X. Luo, and M. Zhou, “Adaptive divergence-based non-negative latent factor analysis of high-dimensional and incomplete matrices from industrial applications,”IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 8, no. 2, pp. 1209–1222, 2024

    2024

  69. [69]

    An adaptively bias-extended non-negative latent factorization of tensors model for accurately representing the dynamic qos data,

    X. Xu, M. Lin, X. Luo, and Z. Xu, “An adaptively bias-extended non-negative latent factorization of tensors model for accurately representing the dynamic qos data,”IEEE Transactions on Services Computing, 2025

    2025

  70. [70]

    A robust coevolutionary neural-based optimization algorithm for constrained nonconvex optimization,

    L. Wei, L. Jin, and X. Luo, “A robust coevolutionary neural-based optimization algorithm for constrained nonconvex optimization,”IEEE Transactions on Neural Networks and Learning Systems, vol. 35, no. 6, pp. 7778–7791, 2022

    2022

  71. [71]

    A scalable multichannel sentiment analysis model with enhanced semantic understanding and redundancy reduction,

    J. Liu, X. Li, M. Lin, and X. Luo, “A scalable multichannel sentiment analysis model with enhanced semantic understanding and redundancy reduction,”IEEE Transactions on Computational Social Systems, 2025

    2025

  72. [72]

    A novel tensor decomposition-based efficient detector for low-altitude aerial objects with knowledge distillation scheme,

    N. Zeng, X. Li, P. Wu, H. Li, and X. Luo, “A novel tensor decomposition-based efficient detector for low-altitude aerial objects with knowledge distillation scheme,”IEEE/CAA Journal of Automatica Sinica, vol. 11, no. 2, pp. 487–501, 2024

    2024

  73. [73]

    A fast nonnegative autoencoder-based approach to latent feature analysis on high-dimensional and incomplete data,

    F. Bi, T. He, and X. Luo, “A fast nonnegative autoencoder-based approach to latent feature analysis on high-dimensional and incomplete data,”IEEE Transactions on Services Computing, vol. 17, no. 3, pp. 733–746, 2023

    2023

  74. [74]

    A sampling-neighborhood-regularized latent factorization of tensor for dynamic qos estimation,

    X. Xu, M. Lin, Z. Xu, and X. Luo, “A sampling-neighborhood-regularized latent factorization of tensor for dynamic qos estimation,”IEEE Transactions on Network and Service Management, vol. 23, pp. 1707–1722, 2025

    2025

  75. [75]

    Attention-mechanism-based neural latent-factorization-of-tensors model,

    ——, “Attention-mechanism-based neural latent-factorization-of-tensors model,”ACM Transactions on Knowledge Dis- covery from Data, vol. 19, no. 4, pp. 1–27, 2025

    2025

  76. [76]

    Graph linear convolution pooling for learning in incomplete high-dimensional data,

    F. Bi, T. He, Y .-S. Ong, and X. Luo, “Graph linear convolution pooling for learning in incomplete high-dimensional data,” IEEE Transactions on Knowledge and Data Engineering, vol. 37, no. 4, pp. 1838–1852, 2024

    2024

  77. [77]

    An outlier-resilient autoencoder for representing high-dimensional and incomplete data,

    D. Wu, Y . Hu, K. Liu, J. Li, X. Wang, S. Deng, N. Zheng, and X. Luo, “An outlier-resilient autoencoder for representing high-dimensional and incomplete data,”IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 9, no. 2, pp. 1379–1391, 2024

    2024

  78. [78]

    Robust low-rank latent feature analysis for spatiotemporal signal recovery,

    D. Wu, Z. Li, Z. Yu, Y . He, and X. Luo, “Robust low-rank latent feature analysis for spatiotemporal signal recovery,” IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 2, pp. 2829–2842, 2023

    2023

  79. [79]

    Neural tucker factorization,

    P. Tang and X. Luo, “Neural tucker factorization,”IEEE/CAA Journal of Automatica Sinica, vol. 12, no. 2, pp. 475–477, 2025

    2025

  80. [80]

    Link-based attributed graph clustering via approximate generative bayesian learning,

    Y . Yang, L. Hu, G. Li, D. Li, P. Hu, and X. Luo, “Link-based attributed graph clustering via approximate generative bayesian learning,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2025

    2025

Showing first 80 references.