Pith. sign in

REVIEW 3 major objections 6 minor 54 references

AttriReBoost: A Gradient-Free Propagation Optimization Method for Cold Start Mitigation in Attribute Missing Graphs

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read AttriReBoost claims to fix the cold start problem in attribute-missing graphs by redefining boundary conditions and adding virtual edges to feature propagation, with reported gains of 5.11% average accuracy and 16-second processing for a…

desk verdict Useful, simple feature-propagation variant with a real but fixable convergence-proof gap and an overstated headline number. read the letter →

arxiv 2501.00743 v1 pith:RUZLDLXZ submitted 2025-01-01 cs.LG cs.AI

Mengran Li , Chaojun Ding , Junzhou Chen , Wenbin Xing , Cong Ye , Ronghui Zhang , Songlin Zhuang , Jia Hu
show 2 more authors
Tony Z. Qiu Huijun Gao
This is my paper · ORCID
classification cs.LGcs.AI
keywords attribute-missinggraphscoldstartproblemfeaturepropagationmissingattributereconstructionvirtualedgesgraphneuralnetworkslarge-scalelearningBanachfixedpointtheorem
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to establish that the cold start problem in attribute-missing graphs can be eliminated within the simple, gradient-free feature propagation paradigm. The proposed method, AttriReBoost (ARB), replaces the hard reset of known nodes after each propagation step with a moving reset that remembers their original features only gradually, and it overlays the graph with a virtual complete-graph term so that every node, including isolated and low-degree ones, receives global mean information. On eight benchmark graphs the authors report better reconstruction and downstream classification than strong prior methods, with the largest gains on isolated and low-degree nodes, and running time nearly equal to feature propagation: the 2.49 million-node Ogbn-Products graph is processed in about 16 seconds. If these results hold, ARB is a drop-in preprocessing step for any GNN needing complete attributes, adding only two hyperparameters over feature propagation and no backpropagation.

What carries the argument

The load-bearing object is the augmented Dirichlet-energy objective $\mathcal{L}(X) = \mathrm{tr}(X^\top L X) + \eta(X_k - Z_k)^\top(X_k - Z_k) + \theta X^\top L_1 X$, where $L$ is the normalized Laplacian and $L_1$ is the Laplacian of the complete graph, so the last term acts as a global-mean penalty. The algorithm is the fixed-point iteration this objective induces: global propagation $X \leftarrow \alpha \tilde{A}X + (1-\alpha)\bar{X}$, followed by moving reset $X_k \leftarrow \beta X_k + (1-\beta)Z_k$. Convergence rests on the iteration matrix $K$ being elementwise dominated by a strongly connected matrix of spectral radius at most 1, which forces $\rho(K) < 1$ and makes the update a Banach contraction.

What would settle it

Run ARB against feature propagation on a synthetic graph with two well-separated communities whose known attributes are far apart, concentrating missing nodes at isolates and degree-one nodes; if ARB's reconstruction error for those nodes is larger than feature propagation's because the global-mean term pulls them toward the wrong average, then the virtual-edge mechanism is the cause.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that missing attributes in a graph are better reconstructed by minimizing a modified Dirichlet energy that contains three terms: the usual edge smoothness, a soft penalty pulling known nodes back toward their original values, and a complete-graph penalty that connects every pair of nodes through virtual edges. The stationarity condition of this objective yields the iteration $X^{(l+1)} = \alpha \tilde{A}X^{(l)} + (1-\alpha)\bar{X}^{(l)}$ for all nodes and $X_k^{(l+1)} = \beta X_k^{(l)} + (1-\beta)Z_k$ for known nodes, where $\alpha$ and $\beta$ are functions of the penalty weights $\theta$ and $\eta$. The paper proves that the iteration matrix has spectral radius strictly below 1, so the update map is a contraction and, by the Banach fixed point theorem, converges to a unique fixed point. Experiments then show faster and more accurate reconstruction than feature propagation and PCFI, especially for isolated and low-degree nodes.

Load-bearing premise

The paper's load-bearing premise is that the three-term loss in its optimization objective, smoothness over real edges, a soft pull toward known attributes, and a complete-graph mean penalty, is the right objective for reconstructing missing attributes; this premise is validated by tuning $\alpha$ and $\beta$ per dataset rather than derived from a generative model of why attributes are missing.

Editorial extensions

If this is right

  • If ARB works as claimed, feature propagation becomes a stronger preprocessing step: any downstream GNN can consume fully imputed attributes without changing the model or adding training cost.
  • Reconstruction quality stops depending so sharply on node degree, so isolated and low-degree nodes, the long tail of real networks, get usable features rather than zero or stale values.
  • Because ARB costs almost the same as feature propagation and avoids backpropagation, graphs with tens of millions of edges can be imputed on a single GPU in seconds, sidestepping memory-heavy batch training.
  • Deeper propagation remains stable up to $l = 10$ in the reported experiments, so users do not need to tune depth carefully to avoid oversmoothing while using ARB.
  • The Banach fixed-point argument means the iteration is guaranteed to reach the same unique steady state from zero initialization, making early stopping a safe speedup rather than a heuristic.

Reading between the lines

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

  • A testable extension is to replace the single global mean with per-community means computed from known nodes; the paper's virtual-edge term assumes one global average, so a clustered-corruption benchmark would reveal whether the mean term blurs community-specific attributes.
  • The moving reset coefficient $\beta$ interpolates between hard reset and pure averaging, so an untested corollary is that annealing $\beta$ from 1 to a lower value over iterations could combine early local fidelity with later global smoothing.
  • ARB's update is a form of diffusion toward a global sink, and the heuristic hyperparameter search over $\alpha$ suggests a direct connection between $\alpha$ and graph conductance; one could predict good $\alpha$ values from the spectral gap instead of searching per dataset.
  • Because the paper's 5.11% average gain is measured under its chosen missing-rate protocol and baseline set, the claim should not be extrapolated to non-random missingness without a dedicated experiment.
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 AttriReBoost (ARB), an iterative, gradient-free method for reconstructing missing node attributes in graphs. ARB augments classic feature propagation with two modifications: a soft "moving reset" of known nodes (weighted by beta) and a virtual complete-graph edge term implementing a global mean (weighted by alpha), derived from a regularized Dirichlet-energy objective. The authors give an iterative update rule, claim convergence via the Banach fixed-point theorem, and present experiments on eight datasets covering reconstruction metrics, downstream node classification, ablations, missing-rate robustness, and runtime. The headline claims are a 5.11% average accuracy improvement over state-of-the-art methods and processing a 2.49M-node graph in about 16 seconds on a single GPU.

Significance. If correct, ARB is an appealing drop-in preprocessing method: it is gradient-free, adds only two hyperparameters relative to feature propagation, scales to millions of nodes, and appears to help precisely in the cold-start regimes (isolated and low-degree nodes) where feature propagation struggles. The paper makes its code public and reports a wide range of benchmarks, which are strengths. However, the convergence proof as written is invalid, the method definition has an ambiguity between Eqs. (9) and (10), and the headline accuracy improvement in the abstract does not match the value reported in Section V-C. These issues affect the paper's central theoretical and empirical claims, though they appear fixable within the manuscript's scope.

major comments (3)
  1. [Section IV-D, Eqs. (11)-(12)] The contraction argument is mathematically incorrect. The iteration matrix K defined in Eq. (10) is not symmetric, so rho(K) < 1 does not imply rho(K^T K) < 1; a simple counterexample is K = [[0,10],[0.01,0]], for which rho(K) = 0.316 but rho(K^T K) = 100. Consequently Eq. (12) does not establish that f is a contraction in the spectral norm, and the Banach Fixed Point Theorem is not legitimately invoked. The same paragraph also writes rho(K) = max_X X^T K X / X^T X, a Rayleigh-quotient identity valid only for symmetric matrices, and the equality ||K Delta X||_2 = Delta X^T K^T K Delta X omits a square on the left-hand side. The 'Additionally' paragraph proposing a block decomposition of B is unexplained and does not apply to the full matrix B. Since the abstract states that convergence is 'rigorously established,' this is a load-bearing gap. A Perron-Frobenius argument on the positive matrix B, with K <= B elementwise and K != B, can likely yield rho(K) < 1 directly, and then convergence of the linear iteration follows from standard spectral-radius theory; the authors should replace the invalid argument with such a proof.
  2. [Section IV-C/D, Eq. (9)-(10), Algorithm 1] The iterative scheme is not defined unambiguously. Eq. (9) writes X_k^{(l+1)} = beta X_k^{(l)} + (1-beta) Z_k, but the fixed-point relation in Eq. (8) and the block iteration in Eq. (10) require X_k^{(l+1)} = beta [alpha \tilde{A} X^{(l)} + (1-alpha) \bar{X}^{(l)}]_k + (1-beta) Z_k. The latter includes the beta B_{ku} X_u term appearing in Eq. (10); the former does not. Algorithm 1 line 4 ('Xk <- beta Xk + (1-beta) Zk') is ambiguous about whether the right-hand side uses the value after line 3. If it does, the algorithm agrees with Eq. (8); if it does not, it defines a different method whose convergence matrix is not the K analyzed in the proof. Please state the order of updates unambiguously and align Eq. (9) with Eq. (10), also correcting the notation so that the global-propagation line uses \bar{X}^{(l)} rather than X^{(l)}.
  3. [Abstract, Section V-C, Table IV] The headline accuracy claim is inconsistent. The abstract and the contributions list say 'average accuracy improvement of 5.11%' over state-of-the-art methods, but Section V-C reports 'an average improvement of 2.49% over the second-best method.' Table IV gives ARB average 79.79% versus PCFI 77.85%, which is a 1.94-percentage-point or 2.49% relative improvement; the 5.11% figure does not follow from any reported table. Please reconcile the numbers and state explicitly whether improvements are reported as relative or absolute percentage points.
minor comments (6)
  1. [Tables II-III] Table II uses 'MEGAE' while the text and Table III use 'MAGAE'; please standardize the method name.
  2. [Table III] Table III uses 'CSAT' for the method called 'CAST' in Section II-A and Table II; please standardize the name.
  3. [Appendix A/B] The supplementary appendix refers to 'Section III D' twice in the convergence proofs; the correct reference is Section IV-D.
  4. [Appendix A, Eq. (18)] The step from X_k = beta (\tilde{A}X)_k + (1-beta) Z_k and X_u = (\tilde{A}X)_u to the displayed 'So: X = \tilde{A}X, X_k = beta X_k + (1-beta) Z_k' is algebraically unclear; please expand or correct the derivation.
  5. [Section V-E] The hyperparameter search reports a range for alpha but not the chosen alpha, beta, and l for each dataset; listing the selected values in a table would improve reproducibility.
  6. [Section IV-D, Eq. (14)] The positive-definiteness argument for L + eta I0_k + theta L1 should explicitly state the assumption that the known set V_k is nonempty or that theta > 0, and that the Rayleigh quotient is taken with the Euclidean norm.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: ARB's derivation follows algebraically from its stated objective, and the reported comparisons are against held-out test labels.

full rationale

The central derivation is self-contained. Equation (9) is obtained by algebraic manipulation of the stated objective in Equation (5), with alpha and beta reparametrizing the arbitrary weights eta and theta; no fitted quantity is renamed as a prediction, since hyperparameters are tuned on a validation split and the reported recall/accuracy figures are computed on test nodes. The self-citations ([3], [9], [37]) appear only in related-work enumerations and as a comparison baseline (CAST/CSAT); none of them supplies a load-bearing premise, uniqueness theorem, or ansatz for the ARB derivation. The convergence proof in Section IV-D does contain a mathematical gap: Eq. (12) infers ||K||_2 < 1 from rho(K) < 1 for a nonsymmetric K, which is not valid in general, and the supplementary appendix refers to a nonexistent 'Section III D'. This is a correctness/rigor problem, not a circular reduction of the result to its inputs. The abstract's 5.11% improvement also lacks direct table support (Table IV supports a 2.49% average gain over the second-best method), another consistency issue that does not change the circularity verdict.

Assumptions & free parameters 3 free parameters · 5 assumptions · 2 invented entities

The central algorithm depends on two ad hoc design choices: a moving reset for known nodes (boundary conditions) and a fully connected virtual-edge overlay (global mean term). These are implemented through hyperparameters α and β that are tuned per dataset on validation data. The mathematical proof of convergence is standard linear algebra and does not add free parameters beyond the loss weights.

free parameters (3)
  • alpha (α) = dataset-specific, ~0.9 to 1.0 (Figure 5)
    Weights original graph propagation vs the global-mean virtual-edge term; tuned via Heuristic Hyperparameter Searcher on validation nDCG/CORR.
  • beta (β) = dataset-specific, no fixed value
    Controls the moving reset strength for known nodes; tuned per dataset.
  • number of iterations l = dataset-specific, up to ~10 in Figure 6
    Propagation depth; tuned per dataset.
assumptions (5)
  • standard math Symmetric normalized adjacency matrix \tilde A has spectral radius at most 1
    Used to bound ρ(B) in Section IV-D; standard spectral graph theory result.
  • standard math If B is a nonnegative irreducible matrix and 0 ≤ K ≤ B with K ≠ B, then ρ(K) < ρ(B)
    Invoked in Section IV-D to show the iteration matrix K has spectral radius below 1.
  • standard math Banach fixed point theorem
    Used to conclude convergence of the contraction mapping X = KX + C.
  • domain assumption The graph is undirected and node attributes are smooth over edges
    Underlies the Dirichlet energy objective inherited from FP (Equation 1).
  • ad hoc to paper The regularized loss in Equation (5) with boundary and virtual-edge penalties is an appropriate objective for attribute reconstruction
    The η and θ weights are chosen ad hoc; the resulting α and β are fitted on validation data.
invented entities (2)
  • Virtual edges (fully connected overlay graph)
    purpose: Connect every node pair to improve propagation to low-degree and isolated nodes via a global mean term.
    The virtual edges are a mathematical construct in the loss (Equation 5) and implementation (global mean term). They have no observable handle outside the paper; the paper itself notes they introduce noise that must be balanced by tuning α.
  • Redefined boundary conditions (moving reset)
    purpose: Adjust known node initialization dynamically during propagation, replacing hard resets with a moving average toward original values.
    This is a design modification to the update rule, not an independently observable construct; its value is demonstrated only through experiments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AttriReBoost: A Gradient-Free Propagation Optimization Method for Cold Start Mitigation in Attribute Missing Graphs." pith.science (2026). https://pith.science/paper/RUZLDLXZ

@misc{pith2026250100743,
  author       = {Pith},
  title        = {Pith review of: AttriReBoost: A Gradient-Free Propagation Optimization Method for Cold Start Mitigation in Attribute Missing Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RUZLDLXZ}},
  note         = {Machine review of arXiv:2501.00743}
}
read the original abstract

Missing attribute issues are prevalent in the graph learning, leading to biased outcomes in Graph Neural Networks (GNNs). Existing methods that rely on feature propagation are prone to cold start problem, particularly when dealing with attribute resetting and low-degree nodes, which hinder effective propagation and convergence. To address these challenges, we propose AttriReBoost (ARB), a novel method that incorporates propagation-based method to mitigate cold start problems in attribute-missing graphs. ARB enhances global feature propagation by redefining initial boundary conditions and strategically integrating virtual edges, thereby improving node connectivity and ensuring more stable and efficient convergence. This method facilitates gradient-free attribute reconstruction with lower computational overhead. The proposed method is theoretically grounded, with its convergence rigorously established. Extensive experiments on several real-world benchmark datasets demonstrate the effectiveness of ARB, achieving an average accuracy improvement of 5.11% over state-of-the-art methods. Additionally, ARB exhibits remarkable computational efficiency, processing a large-scale graph with 2.49 million nodes in just 16 seconds on a single GPU. Our code is available at https://github.com/limengran98/ARB.

Figures

Figures reproduced from arXiv: 2501.00743 by the authors.

Figure 1
Figure 1. (a) Cold start examples in attribute-missing graphs. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall framework of ARB. ARB address the cold [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Specific implementation process of attribute reconstruc [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Training process and convergence speed. accuracy, achieving satisfactory results early (Epoch=2) and enabling early stopping. This rapid convergence reduces overall computation time, making ARB more efficient for large￾scale and time-sensitive applications. Similar to …
Figure 6
Figure 6. Figure 6: Comparison of different propagation times [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 5
Figure 5. Figure 5: Hyperparameter value α and β validation. F. Sensitivity Analysis on Missing Rates (Q5) We conduct a series of attribute reconstruction experiments to validate the robustness of ARB, with missing attribute rates (a) Cora (b) CiteSeer [PITH_FULL_IMAGE:figures/full_fig_p…
Figure 7
Figure 7. Figure 7: Comparison of running time. TABLE X: Comparison of running time (s). Best improvement is blue. Method Cora CiteSeer PubMed Computers FP 0.2327 0.2653 0.2434 0.3152 PCFI 1.1521(×4.95) 1.2196(×4.60) 1.2136(×4.99) 1.2383(×3.93) Ours 0.2351(×1.01) 0.2654(×1.01) 0.2549(×1.0…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 35 canonical work pages

  1. [1]

    Smoothness sensor: Adaptive smoothness-transition graph convolutions for attributed graph clustering,

    C. Ji, H. Chen, R. Wang, Y . Cai, and H. Wu, “Smoothness sensor: Adaptive smoothness-transition graph convolutions for attributed graph clustering,” IEEE Transactions on Cybernetics , vol. 52, no. 12, pp. 12 771–12 784, 2022

    work page 2022

  2. [2]

    Harnessing multimodal large language models for traffic knowledge graph generation and decision-making,

    S. Kuang, Y . Liu, X. Wang, X. Wu, and Y . Wei, “Harnessing multimodal large language models for traffic knowledge graph generation and decision-making,” Communications in Transportation Research , vol. 4, p. 100146, 2024

    work page 2024

  3. [3]

    Redundancy is not what you need: An embedding fusion graph auto-encoder for self-supervised graph representation learning,

    M. Li, Y . Zhang, S. Wang, Y . Hu, and B. Yin, “Redundancy is not what you need: An embedding fusion graph auto-encoder for self-supervised graph representation learning,” IEEE Transactions on Neural Networks and Learning Systems , 2024

    2024

  4. [4]

    Semi-supervised graph structure learning via dual reinforcement of label and prior structure,

    R. Yuan, Y . Tang, Y . Wu, J. Niu, and W. Zhang, “Semi-supervised graph structure learning via dual reinforcement of label and prior structure,” IEEE Transactions on Cybernetics, vol. 54, no. 11, pp. 6943–6956, 2024

    work page 2024

  5. [5]

    Multi-scale Temporal Fusion Transformer for Incomplete Vehicle Trajectory Prediction

    Z. Liu, C. Li, Y . Wang, N. Yang, X. Fan, J. Ma, and X. Zhao, “Multi-scale temporal fusion transformer for incomplete vehicle trajectory prediction,” arXiv preprint arXiv:2409.00904 , 2024

    work page Pith review arXiv 2024

  6. [6]

    Graph representation learning via graphical mutual information maxi- mization,

    Z. Peng, W. Huang, M. Luo, Q. Zheng, Y . Rong, T. Xu, and J. Huang, “Graph representation learning via graphical mutual information maxi- mization,” in Proceedings of The Web Conference , 2020, pp. 259–270

    work page 2020

  7. [7]

    A comprehensive survey on imputation of missing data in internet of things,

    D. Adhikari, W. Jiang, J. Zhan, Z. He, D. B. Rawat, U. Aickelin, and H. A. Khorshidi, “A comprehensive survey on imputation of missing data in internet of things,” ACM Computing Surveys , vol. 55, no. 7, pp. 1–38, 2022

    work page 2022

  8. [8]

    A novel representation learning for dynamic graphs based on graph convolutional networks,

    C. Gao, J. Zhu, F. Zhang, Z. Wang, and X. Li, “A novel representation learning for dynamic graphs based on graph convolutional networks,” IEEE Transactions on Cybernetics , vol. 53, no. 6, pp. 3599–3612, 2023

    work page 2023

Show all 54 references
  1. [9]

    Scae: Struc- tural contrastive auto-encoder for incomplete multi-view representation learning,

    M. Li, R. Zhang, Y . Zhang, X. Piao, S. Zhao, and B. Yin, “Scae: Struc- tural contrastive auto-encoder for incomplete multi-view representation learning,” ACM Transactions on Multimedia Computing, Communications and Applications, vol. 20, no. 9, pp. 1–24, 2024

    2024

  2. [10]

    Handling missing data with graph representation learning,

    J. You, X. Ma, D. Y . Ding, M. Kochenderfer, and J. Leskovec, “Handling missing data with graph representation learning,” in Proceedings of the International Conference on Neural Information Processing Systems , 2020, pp. 19 075–19 087

    2020

  3. [11]

    Fair attribute completion on graph with missing attributes,

    D. Guo, Z. Chu, and S. Li, “Fair attribute completion on graph with missing attributes,” arXiv preprint arXiv:2302.12977 , 2023

    2023 arXiv

  4. [12]

    Learning on attribute-missing graphs,

    X. Chen, S. Chen, J. Yao, H. Zheng, Y . Zhang, and I. W. Tsang, “Learning on attribute-missing graphs,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 2, pp. 740–757, 2022. 11

    2022

  5. [13]

    Accurate node feature estimation with structured variational graph autoencoder,

    J. Yoo, H. Jeon, J. Jung, and U. Kang, “Accurate node feature estimation with structured variational graph autoencoder,” in Proceedings of the ACM SIGKDD Conference on Knowledge Discovery and Data Mining , 2022, pp. 2336–2346

    2022

  6. [14]

    Initializing then refining: A simple graph attribute imputation network,

    W. Tu, S. Zhou, X. Liu, Y . Liu, Z. Cai, E. Zhu, Z. Changwang, and J. Cheng, “Initializing then refining: A simple graph attribute imputation network,” in Proceedings of the International Joint Conference on Artificial Intelligence, 2022, pp. 3494–3500

    2022

  7. [15]

    Measuring and relieving the over-smoothing problem for graph neural networks from the topological view,

    D. Chen, Y . Lin, W. Li, P. Li, J. Zhou, and X. Sun, “Measuring and relieving the over-smoothing problem for graph neural networks from the topological view,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 34, no. 04, 2020, pp. 3438–3445

    2020

  8. [16]

    On the unreasonable effectiveness of feature propagation in learning on graphs with missing node features,

    E. Rossi, H. Kenlay, M. I. Gorinova, B. P. Chamberlain, X. Dong, and M. M. Bronstein, “On the unreasonable effectiveness of feature propagation in learning on graphs with missing node features,” in Learning on Graphs Conference . PMLR, 2022, pp. 11–1

    2022

  9. [17]

    Confidence-based feature imputation for graphs with partially known features,

    D. Um, J. Park, S. Park, and J. Y . Choi, “Confidence-based feature imputation for graphs with partially known features,” arXiv preprint arXiv:2305.16618, 2023

    2023 arXiv

  10. [18]

    Cold brew: Distilling graph node representations with incomplete or missing neighborhoods,

    W. Zheng, E. W. Huang, N. Rao, S. Katariya, Z. Wang, and K. Subbian, “Cold brew: Distilling graph node representations with incomplete or missing neighborhoods,” arXiv preprint arXiv:2111.04840 , 2021

    2021 arXiv

  11. [19]

    Multiview deep graph infomax to achieve unsupervised graph embedding,

    Z. Zhou, Y . Hu, Y . Zhang, J. Chen, and H. Cai, “Multiview deep graph infomax to achieve unsupervised graph embedding,” IEEE Transactions on Cybernetics, vol. 53, no. 10, pp. 6329–6339, 2023

    2023

  12. [20]

    Predict then propagate: Graph neural networks meet personalized pagerank,

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

    2018 arXiv

  13. [21]

    A Stochastic Approximation Method,

    H. Robbins and S. Monro, “A Stochastic Approximation Method,” The Annals of Mathematical Statistics , vol. 22, no. 3, pp. 400 – 407, 1951

    1951

  14. [22]

    Navigating networks by using homophily and degree,

    ¨O. S ¸ims ¸ek and D. Jensen, “Navigating networks by using homophily and degree,” Proceedings of the National Academy of Sciences , vol. 105, no. 35, pp. 12 758–12 762, 2008

    2008

  15. [23]

    A singular value thresholding algorithm for matrix completion,

    J.-F. Cai, E. J. Cand `es, and Z. Shen, “A singular value thresholding algorithm for matrix completion,” SIAM Journal on Optimization, vol. 20, no. 4, pp. 1956–1982, 2010

    1956

  16. [24]

    Incomplete multiview spectral clustering with adaptive graph learning,

    J. Wen, Y . Xu, and H. Liu, “Incomplete multiview spectral clustering with adaptive graph learning,” IEEE Transactions on Cybernetics, vol. 50, no. 4, pp. 1418–1429, 2020

    2020

  17. [25]

    Generalized incomplete multiview clustering with flexible locality structure diffusion,

    J. Wen, Z. Zhang, Z. Zhang, L. Fei, and M. Wang, “Generalized incomplete multiview clustering with flexible locality structure diffusion,” IEEE Transactions on Cybernetics , vol. 51, no. 1, pp. 101–114, 2021

    2021

  18. [26]

    Tensor completion-based incomplete multiview clustering,

    W. Xia, Q. Gao, Q. Wang, and X. Gao, “Tensor completion-based incomplete multiview clustering,” IEEE Transactions on Cybernetics , vol. 52, no. 12, pp. 13 635–13 644, 2022

    2022

  19. [27]

    Differentiated anchor quantity assisted incomplete multiview clustering without number-tuning,

    S. Yu, P. Zhang, S. Wang, Z. Dong, H. Yang, E. Zhu, and X. Liu, “Differentiated anchor quantity assisted incomplete multiview clustering without number-tuning,” IEEE Transactions on Cybernetics , vol. 54, no. 11, pp. 7024–7037, 2024

    2024

  20. [28]

    Gain: Missing data imputation using generative adversarial nets,

    J. Yoon, J. Jordon, and M. Schaar, “Gain: Missing data imputation using generative adversarial nets,” in Proceedings of the International Conference on Machine Learning . PMLR, 2018, pp. 5689–5698

    2018

  21. [29]

    Graph recurrent networks with attributed random walks,

    X. Huang, Q. Song, Y . Li, and X. Hu, “Graph recurrent networks with attributed random walks,” in Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , 2019, pp. 732–740

    2019

  22. [30]

    Attributed random walk as matrix factorization,

    L. Chen, J. Bruna, and M. Bronstein, “Attributed random walk as matrix factorization,” in Proceedings of the International Conference on Neural Information Processing Systems, Graph Representation Learning Workshop, 2019

    2019

  23. [31]

    Missing data imputation with adversarially-trained graph convolutional networks,

    I. Spinelli, S. Scardapane, and A. Uncini, “Missing data imputation with adversarially-trained graph convolutional networks,” Neural Networks, vol. 129, pp. 249–260, 2020

    2020

  24. [32]

    Graph convolutional networks for graphs containing missing features,

    H. Taguchi, X. Liu, and T. Murata, “Graph convolutional networks for graphs containing missing features,” Future Generation Computer Systems, vol. 117, pp. 155–168, 2021

    2021

  25. [33]

    Semi-supervised classification with graph convolutional networks,

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

    2016 arXiv

  26. [34]

    Local computations with proba- bilities on graphical structures and their application to expert systems,

    S. L. Lauritzen and D. J. Spiegelhalter, “Local computations with proba- bilities on graphical structures and their application to expert systems,” Journal of the Royal Statistical Society: Series B (Methodological) , vol. 50, no. 2, pp. 157–194, 1988

    1988

  27. [35]

    Amer: A new attribute-missing network embedding approach,

    D. Jin, R. Wang, T. Wang, D. He, W. Ding, Y . Huang, L. Wang, and W. Pedrycz, “Amer: A new attribute-missing network embedding approach,” IEEE Transactions on Cybernetics , vol. 53, no. 7, pp. 4306– 4319, 2022

    2022

  28. [36]

    Handling missing data via max-entropy regularized graph autoencoder,

    Z. Gao, Y . Niu, J. Cheng, J. Tang, L. Li, T. Xu, P. Zhao, F. Tsung, and J. Li, “Handling missing data via max-entropy regularized graph autoencoder,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 37, no. 6, 2023, pp. 7651–7659

    2023

  29. [37]

    Csat: Contrastive sampling-aggregating transformer for community detection in attribute-missing networks,

    M. Li, Y . Zhang, W. Zhang, S. Zhao, X. Piao, and B. Yin, “Csat: Contrastive sampling-aggregating transformer for community detection in attribute-missing networks,” IEEE Transactions on Computational Social Systems, vol. 11, no. 2, pp. 2277–2290, 2024

    2024

  30. [38]

    Multi-view graph imputation network,

    X. Peng, J. Cheng, X. Tang, B. Zhang, and W. Tu, “Multi-view graph imputation network,” Information Fusion, vol. 102, p. 102024, 2024

    2024

  31. [39]

    Pre-training graph neural networks for cold-start users and items representation,

    B. Hao, J. Zhang, H. Yin, C. Li, and H. Chen, “Pre-training graph neural networks for cold-start users and items representation,” in Proceedings of the ACM International Conference on Web Search and Data Mining , 2021, pp. 265–273

    2021

  32. [40]

    An unified search and recommendation foundation model for cold-start scenario,

    Y . Gong, X. Ding, Y . Su, K. Shen, Z. Liu, and G. Zhang, “An unified search and recommendation foundation model for cold-start scenario,” in Proceedings of the ACM International Conference on Information and Knowledge Management, 2023, pp. 4595–4601

    2023

  33. [41]

    User cold-start recom- mendation via inductive heterogeneous graph neural network,

    D. Cai, S. Qian, Q. Fang, J. Hu, and C. Xu, “User cold-start recom- mendation via inductive heterogeneous graph neural network,” ACM Transactions on Information Systems , vol. 41, no. 3, pp. 1–27, 2023

    2023

  34. [42]

    Extract the knowledge of graph neural net- works and go beyond it: An effective knowledge distillation framework,

    C. Yang, J. Liu, and C. Shi, “Extract the knowledge of graph neural net- works and go beyond it: An effective knowledge distillation framework,” in Proceedings of the Web Conference , 2021, pp. 1227–1237

    2021

  35. [43]

    Aligning distillation for cold-start item recommendation,

    F. Huang, Z. Wang, X. Huang, Y . Qian, Z. Li, and H. Chen, “Aligning distillation for cold-start item recommendation,” in Proceedings of the International ACM SIGIR Conference on Research and Development in Information Retrieval, 2023, pp. 1147–1157

    2023

  36. [44]

    A variational method in image recovery,

    G. Aubert and L. Vese, “A variational method in image recovery,” SIAM Journal on Numerical Analysis , vol. 34, no. 5, pp. 1948–1979, 1997

    1948

  37. [45]

    A survey of shape analysis techniques,

    S. Loncaric, “A survey of shape analysis techniques,” Pattern Recognition, vol. 31, no. 8, pp. 983–1001, 1998

    1998

  38. [46]

    F. R. Chung, Spectral graph theory. American Mathematical Soc., 1997, vol. 92

    1997

  39. [47]

    Sur les op ´erations dans les ensembles abstraits et leur application aux ´equations int ´egrales,

    S. Banach, “Sur les op ´erations dans les ensembles abstraits et leur application aux ´equations int ´egrales,” Fundamenta Mathematicae, vol. 3, no. 1, pp. 133–181, 1922

    1922

  40. [48]

    Graph attention networks,

    P. Veli ˇckovi´c, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y . Bengio, “Graph attention networks,” arXiv preprint arXiv:1710.10903, 2017

    2017 arXiv

  41. [49]

    Simplifying graph convolutional networks,

    F. Wu, A. Souza, T. Zhang, C. Fifty, T. Yu, and K. Weinberger, “Simplifying graph convolutional networks,” in Proceedings of the Conference on International Conference on Machine Learning . PMLR, 2019, pp. 6861–6871

    2019

  42. [50]

    Revisiting semi-supervised learning with graph embeddings,

    Z. Yang, W. Cohen, and R. Salakhudinov, “Revisiting semi-supervised learning with graph embeddings,” in Proceedings of the International Conference on Machine Learning . PMLR, 2016, pp. 40–48

    2016

  43. [51]

    Pitfalls of graph neural network evaluation,

    O. Shchur, M. Mumme, A. Bojchevski, and S. G ¨unnemann, “Pitfalls of graph neural network evaluation,” arXiv preprint arXiv:1811.05868 , 2018

    2018 arXiv

  44. [52]

    Open graph benchmark: Datasets for machine learning on graphs,

    W. Hu, M. Fey, M. Zitnik, Y . Dong, H. Ren, B. Liu, M. Catasta, and J. Leskovec, “Open graph benchmark: Datasets for machine learning on graphs,” Proceedings of the Conference in Neural Information Processing Systems, vol. 33, pp. 22 118–22 133, 2020

    2020

  45. [53]

    Inductive representation learning on large graphs,

    W. L. Hamilton, R. Ying, and J. Leskovec, “Inductive representation learning on large graphs,” in Proceedings of the International Conference on Neural Information Processing Systems , 2017, pp. 1025–1035. Mengran Li is currently pursuing a Ph.D. degree at the Guangdong Key La...

    2017

  46. [2009]

    After finishing his post-doctoral research work at INRIA, Paris, France, in February 2011, he is currently an Associate Professor with Guangdong Key Laboratory of Intelligent Transportation System, School of Intelligent Systems Engineering, Shenzhen Campus of Sun Yat-sen Unive...

    2011

Pith tools

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