Normality Calibration in Semi-supervised Graph Anomaly Detection

Guansong Pang; Guoguo Ai; Guolei Zeng; Hezhe Qiao; Jinsong Guo

REVIEW 3 major objections 4 minor 16 references

Normality Calibration in Semi-supervised Graph Anomaly Detection

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read GraphNC calibrates the normality learned by any semi-supervised graph anomaly detector by aligning student scores to a teacher's distribution and tightening normal representations, cutting false alarms.

desk verdict Useful empirical framework for semi-supervised GAD, but the main regularization loss is defined with a sign error that makes the published method ambiguous and the theory does not fix it. read the letter →

arxiv 2510.02014 v3 pith:JJDI64WO submitted 2025-10-02 cs.LG

Guolei Zeng , Hezhe Qiao , Guoguo Ai , Jinsong Guo , Guansong Pang This is my paper

classification cs.LG

keywords graphanomalydetectionsemi-supervisedlearningnormalitycalibrationscoredistributionalignmentconsistencyregularizationteacher-studentframeworkneuralnetworks

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

Semi-supervised graph anomaly detection learns normality from a small set of labeled normal nodes, and existing methods tend to overfit those examples, producing high false positives. GraphNC is a teacher-student framework that calibrates this learned normality in two spaces: it aligns the student's anomaly scores to the full score distribution of a pre-trained teacher (ScoreDA), and it regularizes the student's node representations by enforcing consistency between original and randomly masked views of labeled normal nodes (NormReg). The paper claims that jointly optimizing these two losses yields more generalized normality and more separable anomaly scores, consistently improving three different types of teacher models and setting a new state of the art on six benchmark datasets.

What carries the argument

ScoreDA (anomaly score distribution alignment) matches the student's predicted anomaly scores to the teacher's scores on all nodes using an MSE loss; NormReg (perturbation-based normality regularization) randomly masks attributes on labeled normal nodes and minimizes the L2 distance between original and augmented GNN representations of those nodes. A two-layer GNN plus an MLP is the student; the teacher is frozen after pre-training.

What would settle it

Take a teacher model trained on the same graph but with a randomly shuffled normal-label set (so its scores are not meaningful) and run GraphNC with that frozen teacher. If the student still improves over the teacher, the gains are not due to score alignment; if it does not improve, the teacher-quality assumption is confirmed. More directly, record the teacher's per-node errors on a held-out set; if gains vanish when the teacher's incorrect scores are concentrated on the unlabeled nodes used by ScoreDA, the method's reliance on those scores is demonstrated.

Watch

Extended reading notes

Core claim

The central claim is that the score distribution of a pre-trained semi-supervised GAD model already encodes correct normality information for most nodes, so a student model can be trained to imitate that distribution via mean-squared-error alignment, while a perturbation-based consistency loss on labeled normal nodes compensates for the teacher's inaccurate scores. The paper shows that this combination shrinks the intra-class variance of normal anomaly scores, which reduces both false positive and false negative rates, and empirically that the framework lifts performance regardless of whether the teacher is reconstruction-based, one-class-based, or anomaly-generation-based.

Load-bearing premise

GraphNC assumes the pre-trained teacher already produces accurate anomaly scores for most nodes; if the teacher is weak or systematically wrong, score alignment propagates those errors and the method's advantage can disappear.

Editorial extensions

If this is right

Applied on top of a strong teacher (GGAD), GraphNC raises average AUROC from 0.7003 to 0.7533 and AUPRC from 0.2745 to 0.3610 across six datasets.
Wrapping weaker teachers (DOMINANT, OCGNN) also produces large gains, e.g., a 36.2% AUROC improvement on Tolokers for DOMINANT.
The theoretical result implies the student's normal-class score variance is strictly smaller than the teacher's, which translates to provably lower false-positive and false-negative probabilities under a sub-Gaussian assumption.
GraphNC is a plug-in: any pre-trained semi-supervised GAD model can serve as the teacher, and stronger teachers yield stronger students.

Reading between the lines

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

The same teacher-student score alignment plus perturbation consistency could be applied to other one-class or positive-only graph learning tasks, such as fraud detection with only known good accounts.
The compactness of normal representations may hurt when the normal class is multimodal; a testable extension would allow multiple prototypes instead of a single compact cluster.
Because NormReg only uses labeled nodes, its benefit may shrink as the labeled set gets very small; the paper's hyperparameter sensitivity on R is a natural place to test this.

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.

Desk Editor's Note

read the letter

First the takeaway: this is a genuinely useful plug-and-play framework for semi-supervised graph anomaly detection, with strong and consistent empirical results. But the paper as written has a sign error in its central regularization term, NormReg, and the theoretical analysis does not resolve it. The reader and stress-test note are right to flag both.

What is new and what works: GraphNC combines teacher score alignment (ScoreDA) with a perturbation-based consistency loss on labeled normal nodes (NormReg) to improve a pre-trained semi-supervised GAD teacher. The idea of calibrating normality in both score and representation space is sensible, and the design is generic: it is tested with three different teacher families. The experiments are the real contribution. Six benchmark datasets, clear gains in AUROC and AUPRC against GGAD and other baselines, and a sensible ablation study showing both components matter. The plug-and-play nature is a practical contribution to a widely used setting.

Soft spots: the sign error is load-bearing. Eq. (3) and Algorithm 1 define L_NormReg as the negative of the mean squared distance between original and masked representations. Since the total loss is minimized, this maximizes that distance—the opposite of the compactness claim in Section 4.3 and Figure 3. Appendix D, Eq. (8), silently redefines the same loss as positive. So the paper does not actually specify which objective is minimized. If the code uses the positive version, the main text and algorithm are wrong; if it uses the negative version, the reported compactness and performance gains cannot come from the published loss. This needs to be resolved before the method can be reproduced.

The theory does not compensate. Theorem 1's proof bounds the variance of the optimal regression E[T|X] in Eq. (7), not the student score variance, and the step from embedding compactness to reduced score variance is asserted. The FPR/FNR argument in Eqs. (12)–(14) conflates conditional distributions: bounding P(S >= tau | y=0) gives you a false positive bound, not a false negative one, and the claimed equivalence with conditioning on y=1 is invalid. I would either fix these arguments or drop the theorem and keep the paper empirical.

The limitation statement is honest and should stay.

Who this is for: researchers in graph anomaly detection who want a simple teacher-student enhancement that works across many settings. They will get value from the empirical recipe, not from the theory. It deserves a serious referee because the empirical claim is important and likely salvageable, but the revision must fix the sign and either repair or remove the theoretical support. I would ask to see the code to confirm which loss is actually optimized.

Referee Report

3 major / 4 minor

Summary. The paper proposes GraphNC, a teacher-student framework for semi-supervised graph anomaly detection. A pretrained semi-supervised GAD model (the teacher) produces anomaly scores for all nodes; a student GNN+MLP is trained to match those scores (ScoreDA) while also minimizing a perturbation-based consistency loss on labeled normal nodes (NormReg), where a random attribute mask creates a second view. The authors claim that joint training makes normal-node representations more compact, separates normal and abnormal score distributions, lowers FPR/FNR, and consistently improves three types of teacher models on six datasets, achieving state-of-the-art results.

Significance. The empirical study is broad: six datasets, three teacher families, ablations, sensitivity scans, and a theoretical appendix. The consistent improvements in AUROC/AUPRC are useful evidence for the approach. However, the manuscript contains a fundamental ambiguity in the definition of the central loss (a sign inconsistency), and the theoretical proof has serious gaps; the empirical claims cannot be fully evaluated until the objective actually optimized is clarified and the proof is repaired.

major comments (3)

[§4.3, Eq. (3); §4.4 Eq. (4); Algorithm 1 line 12; App. D Eq. (8)] L_NormReg is defined with a leading minus sign, while the total loss is minimized; thus the term maximizes the distance between original and masked representations, contradicting the compactness rationale in §4.3 and Figure 3. Appendix D Eq. (8) silently drops the minus sign. The paper therefore does not define a single method. Please state which loss was actually implemented, correct Eq. (3) and Algorithm 1, and either rerun experiments or show that the sign error is typographical only.
[App. D, Eqs. (6)-(10) and subsequent paragraph] The proof that NormReg shrinks score variance is not established. Eq. (7) yields Var(E[T|X]) <= Var(T), not strict <; no argument connects embedding consistency to score variance. Minimizing E||epsilon - epsilon'||^2 only aligns the two views; it does not force either view to concentrate around mu0, so the claim that population variance around mu0 decreases is a non sequitur. The step Var(F*_S(H)) < Var(F*_S(X)) is asserted without proof. Please either supply a correct derivation or rephrase Theorem 1 as a conjecture/empirical observation.
[App. D, Eqs. (11)-(15)] The FPR/FNR conclusion does not follow. Eq. (12) gives upper bounds; the fact that one bound is smaller does not imply the corresponding probability is smaller. Moreover, replacing P(S_s<tau|y=0) with P(S_s<tau|y=1) is invalid: with binary labels, P(S<tau|y=1) is not the complement of P(S>=tau|y=0). As a result, Theorem 1's claims about reducing FPR and FNR are unproven.

minor comments (4)

[§5.2 and Table 2] The cross-reference says 'results are shown in Table 3'; the teacher-model results are in Table 2, while Table 3 contains the ablation study.
[Appendix C.1 and Fig. 4] Appendix C.1 title 'SORE DISTRIBUTION' should be 'SCORE DISTRIBUTION'; Fig. 4 axis label 'AURPC' should be 'AUPRC'.
[Algorithm 1] Line 2 says 'parameters Theta' for the student model, but Theta is also used for the teacher; line 12 has 'NormaReg' typo; line 15 'Minizing' should be 'Minimizing'.
[Table 1] The paper reports averages over 5 independent runs but no standard deviations. Adding error bars or a variance table would improve the empirical support.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: teacher-student score alignment plus a consistency regularizer is a self-contained objective; evaluations are against ground-truth labels.

full rationale

GraphNC's derivation chain is not circular. ScoreDA (Eq. 2) is an MSE distillation loss that regresses student scores onto the frozen teacher's scores; this is a standard student-teacher setup whose objective is defined independently of the evaluation metric (AUROC/AUPRC on ground-truth labels). NormReg (Eq. 3/Algorithm 1 line 12, and the corrected positive form in Eq. 8) is a consistency regularizer on labeled nodes; its claimed benefit—compact normal representations—is an empirical property tested by ablations and t-SNE, not a quantity that is fitted to the final results. The teacher GGAD is prior work by overlapping authors (Qiao et al., 2024), but it is an externally published, benchmark-evaluated model, not a private uniqueness theorem or ansatz, and the paper's improvement over GGAD is measured against ground truth rather than guaranteed by the distillation objective. The stated limitation that the teacher must be informative is an explicit assumption, not a hidden circular input. The main manuscript does contain a sign inconsistency in Eq. (3) versus Eq. (8), and the proof of Theorem 1 has a gap in moving from embedding compactness to score variance, but these are correctness/rigor concerns, not cases where the prediction reduces to the input by construction. No circular step is present.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The method introduces no new entities. It relies on standard teacher-student distillation and consistency regularization. The main free parameters are hyperparameters, and the key assumptions are about teacher quality and the theoretical connection between representation consistency and score variance.

free parameters (3)

alpha = 0.01
Trade-off weight for NormReg in the total loss; chosen by hand across all datasets, not fitted. The paper reports some sensitivity but no per-dataset tuning.
omega = 0.30
Mask ratio for the attribute masking augmentation in NormReg; set globally to 0.3 after observing dataset-dependent sensitivity.
learning rate = 5e-3 (Photo, Reddit) / 5e-4 (others)
Two learning rates chosen based on attribute dimensions to avoid overfitting; a hand-set schedule affecting the optimization and final results.

assumptions (4)

domain assumption The teacher model provides informative anomaly scores for most nodes, so MSE alignment improves student scores.
The whole ScoreDA component relies on this premise. The authors state it as a limitation and provide no quantitative check of teacher reliability per node.
ad hoc to paper Anomaly scores of the normal class follow a sub-Gaussian distribution.
Used in Appendix D to prove tail bounds without justification from data.
ad hoc to paper Perturbations (epsilon) across the two views are independent and zero-mean.
Critical for deriving the NormReg population variance relation in Appendix D; the masking augmentation does not guarantee independence of these perturbation errors.
ad hoc to paper Minimizing NormReg reduces the score variance of the normal class.
The proof asserts this connection without proof; consistency between two views of each node does not necessarily reduce the dispersion of representations across different normal nodes.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Normality Calibration in Semi-supervised Graph Anomaly Detection." pith.science (2026). https://pith.science/paper/JJDI64WO

@misc{pith2026251002014,
  author       = {Pith},
  title        = {Pith review of: Normality Calibration in Semi-supervised Graph Anomaly Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JJDI64WO}},
  note         = {Machine review of arXiv:2510.02014}
}

read the original abstract

Graph anomaly detection (GAD) has attracted growing interest for its crucial ability to uncover irregular patterns in broad applications. Semi-supervised GAD, which assumes a subset of annotated normal nodes available during training, is among the most widely explored application settings. However, the normality learned by existing semi-supervised GAD methods is limited to the labeled normal nodes, often inclining to overfitting the given patterns. These can lead to high detection errors, such as high false positives. To overcome this limitation, we propose GraphNC , a graph normality calibration framework that leverages both labeled and unlabeled data to calibrate the normality from a teacher model (a pre-trained semi-supervised GAD model) jointly in anomaly score and node representation spaces. GraphNC includes two main components, anomaly score distribution alignment (ScoreDA) and perturbation-based normality regularization (NormReg). ScoreDA optimizes the anomaly scores of our model by aligning them with the score distribution yielded by the teacher model. Due to accurate scores in most of the normal nodes and part of the anomaly nodes in the teacher model, the score alignment effectively pulls the anomaly scores of the normal and abnormal classes toward the two ends, resulting in more separable anomaly scores. Nevertheless, there are inaccurate scores from the teacher model. To mitigate the misleading by these scores, NormReg is designed to regularize the graph normality in the representation space, making the representations of normal nodes more compact by minimizing a perturbation-guided consistency loss solely on the labeled nodes.

Figures

Figures reproduced from arXiv: 2510.02014 by the authors.

**Figure 1.** (a) False positive rate and (b) false negative rate results on Amazon (Dou et al., 2020) and Tolokers (McAuley et al., 2015). (c), (d), and (e) show the score distributions of normal and abnormal nodes for GGAD, ScoreDA, and ScoreDA+NormReg (i.e., GraphNC) on Amazon, where GGAD is used as a teacher model in both ScoreDA and GraphNC. GraphNC includes two main components, namely anomaly score distribution alignment (… view at source ↗

**Figure 2.** Overview of GraphNC. The input graph consists of a small labeled normal node set and [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** (a) and (b) provide t-SNE visualization of the node representations for GraphNC with/without using NormReg. (c) The average deviation of the normal class on Tolokers. normal nodes, simulating diverse normal patterns that may be different from the ones derived directly from the labeled nodes. Thus, enforcing consistency over the representations of these augmented nodes and the labeled normal nodes helps cluster … view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: (a-d) AUROC and AURPC results w.r.t α and ω. (e-f) AUROC results w.r.t R on Photo and Reddit. We evaluate the sensitivity of GraphNC w.r.t. the loss moderator α, mask ratio ω, and training size R. A detailed analysis on time complexity can be found App. E. Performanc…

**Figure 5.** Figure 5: The score distribution of DOMINANT along the corresponding NomrDR enabled DOM [PITH_FULL_IMAGE:figures/full_fig_p014_5.png]

**Figure 6.** Figure 6: The score distribution of OCGNN along the corresponding GraphNC enabled OCGNN [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

16 extracted references · 7 linked inside Pith

[1]

A survey on visual anomaly detection: Challenge, approach, and prospect.arXiv preprint arXiv:2401.16402,

Yunkang Cao, Xiaohao Xu, Jiangning Zhang, Yuqi Cheng, Xiaonan Huang, Guansong Pang, and Weiming Shen. A survey on visual anomaly detection: Challenge, approach, and prospect.arXiv preprint arXiv:2401.16402,

arXiv
[3]

Semi-supervised and un- supervised deep visual learning: A survey.IEEE transactions on pattern analysis and machine intelligence, 46(3):1327–1347, 2022b

Yanbei Chen, Massimiliano Mancini, Xiatian Zhu, and Zeynep Akata. Semi-supervised and un- supervised deep visual learning: A survey.IEEE transactions on pattern analysis and machine intelligence, 46(3):1327–1347, 2022b. Zhenxing Chen, Bo Liu, Meiqing Wang, Peng Dai, Jun Lv, and Liefeng Bo. Generative adversarial attributed network anomaly detection. InPro...

1989
[4]

• Amazon (Dou et al., 2020): It is a co-review network obtained from the Musical Instrument category on Amazon.com

A detailed introduction of these datasets is given as follows. • Amazon (Dou et al., 2020): It is a co-review network obtained from the Musical Instrument category on Amazon.com. There are also three relations: U-P-U (users reviewing at least one same product), U-S-U (users having at least one same star rating within one week), and U-V-U (users with top-5...

2020
[6]

Ad- dressing heterophily in graph anomaly detection: A perspective of graph spectrum

Yuan Gao, Xiang Wang, Xiangnan He, Zhenguang Liu, Huamin Feng, and Yongdong Zhang. Ad- dressing heterophily in graph anomaly detection: A perspective of graph spectrum. InProceed- ings of the ACM Web Conference 2023, pp. 1528–1538, 2023a. Yuan Gao, Xiang Wang, Xiangnan He, Zhenguang Liu, Huamin Feng, and Yongdong Zhang. Al- leviating structural distributi...

2023
[7]

Interlude: Interactions between la- beled and unlabeled data to enhance semi-supervised learning.arXiv preprint arXiv:2403.10658,

Zhe Huang, Xiaowei Yu, Dajiang Zhu, and Michael C Hughes. Interlude: Interactions between la- beled and unlabeled data to enhance semi-supervised learning.arXiv preprint arXiv:2403.10658,

arXiv
[8]

A review of pseudo- labeling for computer vision.arXiv preprint arXiv:2408.07221,

Patrick Kage, Jay C Rothenberger, Pavlos Andreadis, and Dimitrios I Diochnos. A review of pseudo- labeling for computer vision.arXiv preprint arXiv:2408.07221,

arXiv
[12]

Pitfalls of graph neural network evaluation.arXiv preprint arXiv:1811.05868,

Oleksandr Shchur, Maximilian Mumme, Aleksandar Bojchevski, and Stephan G¨unnemann. Pitfalls of graph neural network evaluation.arXiv preprint arXiv:1811.05868,

arXiv
[15]

The local node affinity is calculated on the learned representations on the truncated graphs

directly obtained based on the embedding, it is optimized on truncated graphs where non- homophily edges are removed iteratively. The local node affinity is calculated on the learned representations on the truncated graphs. • GGAD (Qiao et al., 2024): It employs two priors related to anomalies, asymmetric local affinity and egocentric closeness, to genera...

2024

Show all 16 references

[16]

Metric Method Amazon T-Finance Reddit YelpChi Tolokers Photo Avg

Table 5: AUROC and AUPRC results comparison of the proposed method GraphNC and its variants. Metric Method Amazon T-Finance Reddit YelpChi Tolokers Photo Avg. AUROC DOMINANT OT+ScoreDA+NormReg0.8936 0.79850.5260 0.6543 0.6977 0.6829 0.7088 OT+ScoreDA 0.7103 0.7069 0.4623 0.512...
[2016]

Pick and choose: a gnn-based imbalanced learning approach for fraud detection

Yang Liu, Xiang Ao, Zidi Qin, Jianfeng Chi, Jinghua Feng, Hao Yang, and Qing He. Pick and choose: a gnn-based imbalanced learning approach for fraud detection. InProceedings of the web conference 2021, pp. 3168–3177, 2021a. Yixin Liu, Zhao Li, Shirui Pan, Chen Gong, Chuan Zhou...

2021
[2019]

Temporal ensembling for semi-supervised learning.arXiv preprint arXiv:1610.02242,

Samuli Laine and Timo Aila. Temporal ensembling for semi-supervised learning.arXiv preprint arXiv:1610.02242,

arXiv
[2020]

Deep anomaly detection on attributed networks

Kaize Ding, Jundong Li, Rohit Bhanushali, and Huan Liu. Deep anomaly detection on attributed networks. InProceedings of the 2019 SIAM international conference on data mining, pp. 594–

2019
[2021]

Deep learning for video anomaly detection: A review.arXiv preprint arXiv:2409.05383,

Peng Wu, Chengyu Pan, Yuting Yan, Guansong Pang, Peng Wang, and Yanning Zhang. Deep learning for video anomaly detection: A review.arXiv preprint arXiv:2409.05383,

arXiv
[2023]

Anomalydae: Dual autoencoder for anomaly detec- tion on attributed networks

10 Haoyi Fan, Fengbin Zhang, and Zuoyong Li. Anomalydae: Dual autoencoder for anomaly detec- tion on attributed networks. InICASSP 2020-2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 5685–5689. IEEE,

2020
[2024]

Gccad: Graph contrastive coding for anomaly detection.IEEE Trans- actions on Knowledge and Data Engineering, 35(8):8037–8051, 2022a

Bo Chen, Jing Zhang, Xiaokang Zhang, Yuxiao Dong, Jian Song, Peng Zhang, Kaibo Xu, Evgeny Kharlamov, and Jie Tang. Gccad: Graph contrastive coding for anomaly detection.IEEE Trans- actions on Knowledge and Data Engineering, 35(8):8037–8051, 2022a. Hao Chen, Ran Tao, Yue Fan, Y...

arXiv
[2025]

Deep semi-supervised anomaly detection.arXiv preprint arXiv:1906.02694,

Lukas Ruff, Robert A Vandermeulen, Nico G ¨ornitz, Alexander Binder, Emmanuel M ¨uller, Klaus- Robert M ¨uller, and Marius Kloft. Deep semi-supervised anomaly detection.arXiv preprint arXiv:1906.02694,

1906 arXiv

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [1]

A survey on visual anomaly detection: Challenge, approach, and prospect.arXiv preprint arXiv:2401.16402,

Yunkang Cao, Xiaohao Xu, Jiangning Zhang, Yuqi Cheng, Xiaonan Huang, Guansong Pang, and Weiming Shen. A survey on visual anomaly detection: Challenge, approach, and prospect.arXiv preprint arXiv:2401.16402,

arXiv

[2] [3]

Semi-supervised and un- supervised deep visual learning: A survey.IEEE transactions on pattern analysis and machine intelligence, 46(3):1327–1347, 2022b

Yanbei Chen, Massimiliano Mancini, Xiatian Zhu, and Zeynep Akata. Semi-supervised and un- supervised deep visual learning: A survey.IEEE transactions on pattern analysis and machine intelligence, 46(3):1327–1347, 2022b. Zhenxing Chen, Bo Liu, Meiqing Wang, Peng Dai, Jun Lv, and Liefeng Bo. Generative adversarial attributed network anomaly detection. InPro...

1989

[3] [4]

• Amazon (Dou et al., 2020): It is a co-review network obtained from the Musical Instrument category on Amazon.com

A detailed introduction of these datasets is given as follows. • Amazon (Dou et al., 2020): It is a co-review network obtained from the Musical Instrument category on Amazon.com. There are also three relations: U-P-U (users reviewing at least one same product), U-S-U (users having at least one same star rating within one week), and U-V-U (users with top-5...

2020

[4] [6]

Ad- dressing heterophily in graph anomaly detection: A perspective of graph spectrum

Yuan Gao, Xiang Wang, Xiangnan He, Zhenguang Liu, Huamin Feng, and Yongdong Zhang. Ad- dressing heterophily in graph anomaly detection: A perspective of graph spectrum. InProceed- ings of the ACM Web Conference 2023, pp. 1528–1538, 2023a. Yuan Gao, Xiang Wang, Xiangnan He, Zhenguang Liu, Huamin Feng, and Yongdong Zhang. Al- leviating structural distributi...

2023

[5] [7]

Interlude: Interactions between la- beled and unlabeled data to enhance semi-supervised learning.arXiv preprint arXiv:2403.10658,

Zhe Huang, Xiaowei Yu, Dajiang Zhu, and Michael C Hughes. Interlude: Interactions between la- beled and unlabeled data to enhance semi-supervised learning.arXiv preprint arXiv:2403.10658,

arXiv

[6] [8]

A review of pseudo- labeling for computer vision.arXiv preprint arXiv:2408.07221,

Patrick Kage, Jay C Rothenberger, Pavlos Andreadis, and Dimitrios I Diochnos. A review of pseudo- labeling for computer vision.arXiv preprint arXiv:2408.07221,

arXiv

[7] [12]

Pitfalls of graph neural network evaluation.arXiv preprint arXiv:1811.05868,

Oleksandr Shchur, Maximilian Mumme, Aleksandar Bojchevski, and Stephan G¨unnemann. Pitfalls of graph neural network evaluation.arXiv preprint arXiv:1811.05868,

arXiv

[8] [15]

The local node affinity is calculated on the learned representations on the truncated graphs

directly obtained based on the embedding, it is optimized on truncated graphs where non- homophily edges are removed iteratively. The local node affinity is calculated on the learned representations on the truncated graphs. • GGAD (Qiao et al., 2024): It employs two priors related to anomalies, asymmetric local affinity and egocentric closeness, to genera...

2024